I saw a Last Spell guide and decided to click the video to leave a thumbs up just to support you because I have learned so, so much from your previous guides on the game. I’m not math inclined and had no interest in a 42 minute video about numbers. Aside from skipping over some of the intro section, I watched the entire thing and was not bored for a moment. It didn’t feel like 42 minutes and was engaging the entire time. It was also explained in a way that was easily understandable. I can’t wait for an expansion of some sort for this game. It is so fun and manages to scratch several gaming itches at once in a way no other can. Strategic turn based combat, rpg leveling and skills, arpg character loot, base and economy building, and tower defense all rolled into one uniquely engrossing game. Thanks again for all the work you put into your videos. Listening to someone talk about a topic they’re extremely knowledgeable in is like chocolate for my brain, especially when it is something I’m interested in. I’ll once again say that the Last Spell subreddit should just be one post with a link to your RUclips channel.
Incredibly kind words and high praise Carl! Thank you! As for me I'll keep in mind that math based titles can definitely be a serious turn off for folks 😂 I had an amazing time making this video and am looking forward to getting back into the swing of things and making more before said expansion drops (can't wait)! Again thank you and cheers!
Thank you Zwerg! It was a fun project and I'm psyched that I managed to kick it out the door, now I just need to get back into a routine and pump out more content like this to counterbalance my new approach to the sub box which seems to be providing a minor haircut in views :)
It's a pretty spectacular game, perhaps a few bugs to iron out and more content would be quite amazing, but I've enjoyed every bit of the ride! Cheers!
It's a wonderful game if sometimes a little hard to approach. It always means a lot to me to hear that my videos made folk's experiences that much more fun or rewarding!
Very good video, great to see that this beautiful game still has passionate people playing and mathing in it! One thing I always observed is the one thing you said about the omen of dexterity: that it is basically the best omen, I have always wondered why that is the case, whether I was at my early stages in playing and didn't like accuracy level ups or now when I take them mostly if it is blue but chose damage/crit/ect upgrades if not blue and it hit me when I was watching the video, overwhelmingly strong when needed and useless when not was very well put because the truth of this game is that the first 12 accuracy are overwhelmingly strong, so much so that 1 point is a steal for an omen that gives accuracy In another way that I now think it, the accuracy omen allows you to spec in damage immediately which is exactly what we want in this game as the enemies get tougher very quick on high apocalypse levels The reason I bring accuracy up is because I had a realization about reliability which I would like to see yours and others take on it I'll preface by saying that your rant about reliability was very funny and enlightening, please do more rants on stats in the future! My main take, just like accuracy reliability is overwhelmingly strong when needed and useless when not Now please do bring the pitchforks yet, there may be a method to my madness Here is the thing, what do we need to do in this game: to kill enemies, and the best way to kill enemies is in one hit( ie one skill use ) And how do we kill enemies in one hit, by making sure that we roll damage in our damage range that is above the monster's effective hp( hp/armor/block/total ect ) Reliability does just that, increases the minimum damage just enough to kill enemies in one hit Not related to reliability, but the reason why HC is so monstrous becuse it can kill 7 enemies in one AP use and even refund the AP it just used! Average damage on the other hand, reliability's nemesis, is only valuable against bosses and perhaps elites in early nights I've never found any other enemies that I need to do more than 3 skill uses Big enemies usually die easily to momentum/multi-hit skills or one rifle shot, although bosses usually also die to one rifle shot so I should probably not include that in the comparisons If when we have reliability and we don't kill in one hit, reliability is basically useless, every other damage stat is almost always better Also I should say that reliability is good only for 5 weapons because of damage ranges: 1h axe, 2h axe, the two bows and tome, it may be okay on 2h sword and wand 2h axe with the shortbow are probable the best with respect to reliability, level 5 has 121-236, damage range of 115 which is almost equal to its min damage of 121 i.e 10% reliability is almost the same as 10% damage for its min damage, that is very high to close in on its one hit potential shortbow level 5 96-192: exact 96 damage range, same reasoning as the 2h axe applies Since the damage ranges of weapons are the biggest in later rounds, reliability can be looked at as the late game accuracy for some weapons( accuracy of damage more specifically ) and probably the biggest reason reliability is so poor in comparison to other stats late game you will have gear that just gives it on the side, and since reliability has the biggest diminishing returns taking it as a level upgrade is less lucrative And when in late game it is very likely that you no longer need the extra damage as you already have enough damage for one hits Accuracy for instance always rolls in its full range 0 - 100 ( or 5 - 100 accounting the dodge rules ), does not matter how much you have of it Reliability on the other hand does not follow that rule, or if someone has observed they can keen me in on how reliability rows are calculated But in essence with 0 reliability you have 0 - 100% of damage range difference, in the case of the shortbow it would be from 0 to 96 extra damage a 20% increase makes it from 20 - 100%, however now you are rolling 80 values instead of the full 100, i.e reliability is now 80% effective translated in another way 20% reliability is 16% if this makes any sense 50% will have rolls from 50 - 100, now only 50 values to roll, no such problem for accuracy which is why it is good If for instance 20% allowed rolls from 20 to 120 and everything above 100 converted into max damage now it would be much much better imagine that now 50% reliability for instance allows max damage rolls 50% of the time, how good does it sound now? All in all I really wish reliability was better then it is, however it seems to me that for a few weapons it is actually very good and a good rule I have is to have somewhere around 20% before it starts having massive diminishing returns, usually look to have it from items
Omen of dexterity helps both in the early and late game. In the early game, you have a limited amount of power so missing a shot feels particularly awful. In the late game, every hero got to pick something other than accuracy on level-up at least once because you had omen of dexterity. Hard to quantify exactly how much power that is, but it's noticeable. If you are not doing corpse piles, it doesn't help as much early game since you just rush the enemy and kill as fast as possible, but corpse piles are so good that not doing them is running the game on super-hard mode. When you have to just sit around waiting for the enemy to arrive you can't afford misses when the enemy finally enters the haven. (yes, I soften them up as I can, but I have AP remaining on each of the first few turns waiting for enemies to be where I want to actually kill them.)
So I very very rarely run omen of dexterity, especially early the only enemies that have any dodge that is worth worrying about are dodgers, and it's only on ranged damage weapons that accuracy is a significant issue, I think it's also more okay than you think to miss like 4 killing shots in a row and waste 4 AP, it happens to me a lot, and is very very annoying, but this game rarely requires perfection and even more rarely rewards you for it. Reliability really does feel like a very dissapointing stat, the greatest value I can think of it providing is that it can increase the chance of you putting an enemy into a badly wounded state rather than just a wounded state without increasing your chance to kill, but the situations in which that's important are very very slim. Reliability is also pretty abundent from perks as well as gear, I often take cherry picking at tier 1, an interesting wrinkle to that is that Longer weapons, a staple perk, gives you negative reliability, but reliability is capped at 0%, so if you never take reliability at level up or on gear, then Longer weapons has absolutely no downside! The thing I brought up with alden when helping with this project, which is something you were very close to discussing here, is that the thing we actually care about isn't max or average damage, but %chance to kill, and I would be very interested to see how reliability fares in that, because things that increase your average/max damage also increase your overkill damage, but if less than max damage kills, than reliability has a greater comparative value, the issue with doing that is quantifying it would require simulating full night enemy compositions including their stat scaling and looking at what skills kill them in how many uses, which is so so so much more work that at the moment it's just not feasible
Hey @florfall ! So first off expect stat by stat rants coming very soon... and i apologize in advance for them because they will all likely be a little unhinged. So I think the thing about reliability that really broke my heart more than anything was nearly exactly what you're describing. In my mind reliability *should* be the best stat for ensuring that something you *can one-shot* you will one-shot, and sadly it's not even good at that. The reality is that you average specific damage upgrade is exactly as good as reliability at raising minimum damage as reliability... and this is when considering the absolute best case scenario, weapons with a very high damage range. Being able to put number to this phenomenon was the ah-hah moment that really soured me on Reliability entirely. If you can get more minimum, maximum and average damage from one stat and the other provides no maximum and less average damage while only providing the same minimum damage is basically a dead stat in my eyes. To speak to the specific math that you've shared... I think I got lost a bit, which is not to say you don't have a point. The way I conceptualize Reliability is to simply imagine a graph with two lines one charting min damage and the other charting max damage, reliability simply creates a diagonal line between the two shaving the bottom of the region off as min damage converged on max damage. In practice the difference between min and max damage never really get's the extreme and the total distance that it's closing is relatively small. But here's my idea... and this one is going to take some doing so it won't happen overnight. My plan is to write a modded game file that doesn't change the game at all. Instead it just records every hit and notes the ones that could have been impacted by reliability, aka when could a 1-shot have been possible with reliability but was missed. This would illuminate exactly how much reliability could actually be a difference maker. My hypothesis is that maybe around night 3-4 there is a spike where you aren't yet doing "excessive" damage and you'll get a slightly elevated rate of "could have killed" happening and then that too will die down. So we'll see when I make that and who will test it with me! :)
I generally agree Tyson, I think that Omen of Dex is the absolute goat because it really saves you two levels of accuracy to get to what I consider to be the "absolute minimum" you need to function in the mid game. For 1 omen point I'm getting the freedom to pick two level-ups that are tailored to the hero, because end of the say every hero need accuracy.
On the topic of accuracy @abigailmcdowell4248 ... my thinking is pretty simply -- how often will I not take an AP level-up. Answer: I will always take one. So conversely missing a hit is often like losing one AP. Every hero needs accuracy and 12% accuracy is really in the sweet spot to remove 95% of misses in the early game when AP is most at a premium. Having played with the spreadsheet, even enemies having 5% dodge causes accuracy to be the best stat by far... so I think that really has cemented in my mind that Omen of Dex is just king of the castle :) You should definitely try it, the opportunity cost is so low and the value is off the freaking charts :)
An interesting thing about the local maxima problem; if you know what level you are going to end up (for example buying a hero from the inn on the final day) it would be useful to find the true maxima for that given level, because there is no future you need to plan for
So I think my thinking still holds... if you consider a truly extreme example you could imagine a damage field where the hero is worthless until the final level and then BAM they become super human because the slope is so steep on that last level. Obviously with the current dynamics that's not possible... but the idea of ending up slightly stronger for a weaker ascent is still probably not worth it. We want to be winning at every level and rarely is the final level destination worth the sacrifice in the midgame.
I saw in a previous video you said something about the future for the game not looking good as far as updates. But I also read a supposed developer comment on reddit about working on a dlc. So hoping that is happening! Just discovered the game and am in love.
DLC was announced by the devs in the discord! it's just it's been quite a long time since there was a new patch, and there have been far fewer balance changes since releaase than a lot of other games in this space get
Hi @someguy8070 ! Like Abigail said the dev have promised DLC this year, so we're all eagerly anticipating more content and changes. The developers have been quite busy with a few other releases and for a while The Last Spell was feeling a little neglected (there are a few minor bugs here and that that I'd love to see patched), but I think the future of the game is bright assuming the devs deliver on their promises! In other news we (the larger community on discord) have recently been breathing some new life into the game with modding that's been a lot of fun as well! It's a good game to fall in love with, you're in good company :)
So happy to hear they help! I do think that exactly what decisions we make and how they contribute to damage can be a pretty mysterious relationship. Hope this helps and glad you're enjoying the vids!
@9237 so AP is almost always good with the diminishing return coming around 10AP... as for poison - it's a stand alone stat that doesn't scale in concert with anything else, so if you're using poison more is better and at 400% poison it's surprisingly effective without being quite as strong as some of the other options.
@@aldenpotamus And of course you always run the omen of "please, please can I have something other than reliability, please, please with sugar on top" 🤣
Blue reliability is 20% and is easily better than what is typically offered against it. green / beige reliability is generally worse than anything offered against it. Obviously if the blue reliability is up against say, +1 multi-hit, well. I usually *don't* run the "improved level-up odds" perk which makes reliability a lot better than it is if you have the improved odds on re-roll. Alden *always* runs improved level-up so reliability falls off pretty fast.
so 20% reliability is often not that much aditional damage? the change in the lowest possible damage can be fairly significant, but the change in average damage isn't that much, the spreadsheet and calculations that Alden is doing does take into account that you are getting 20% reliability vs 8% physical damage for example, and even with that reliability is never fantastic, I do think that one of the things most lacking from aldens spreadsheet is some expected value of a reroll. I do think that when you look at how the damage numbers shown in the spreadsheet change when you look at a reliabilty vs and opportunism or isolation level up it's pretty extreme, and then you need to consider that green level ups are closer to blue level ups than grey, so I would still broadly agree with Alden that a blue reliability is not great compared to the other stats that improve your damage, I do think there is something to be said for the improved consistency that reliability provides, but it's very hard to assign that a worth. Then a thing that I am hoping to look into with aldens help soon is how much of each stat you can expect to be given from the random attributes that appears on stats, and reliability is one of the stats that gets the biggest numbers from gear, and so recieves even greater diminishing returns from level ups (the biggest example of that for me is the cape that gives 33% reliability at level 5 with no rarity, getting you 1/3 of the way to the cap in one item is huge and massively devalues getting that resource from other places) Alden is very very aware of how contentious Reliability is, and I'm expecting it to get a lot of the attention in the upcoming videos that dive deeper into how the spreadsheet works
Hey Tyson! As always excited to see a comment from you, super appreciate you thoughts on these topics! I've been giving some thought to the hypothesis that advanced training (and my regular use of it) might be coloring my opinion on Reliability and I'm not so sure that it is. The analysis that I did generally considered the highest level of upgrade (simply because "more of the best thing" will always be better than "less of the same thing"). As I mentioned in the video precents are precents when it comes to the value of stat upgrades with the exception of reliability. Since that 20% is operating only on a portion of total damage (specifically the delta between min and max) it doesn't have nearly the magnitude of other stats like physical damage or opportunism where 1% is indeed (ignoring diminishing returns) 1% improvement of your overall damage. I think the thing that was most damning about my reliability analysis was that a comparable rarity improvement of reliability or physical damage doesn't yield more min damage let alone more average damage. As a result it's nearly always better to choose something that will improve your min and max damage as you get twice the impact on average damage and seemingly also a proportional increase to min damage. All this being said, I spent a lot of time in the sheet trying to find places were Reliability was the top dog and while they exist they're pretty rare when you consider a 12-ish level cap... so for now I'll happily take reliability on solid gear pieces but also likely not take it from level-ups except in the most extreme of cases. As always super appreciate the comment! Cheers!
@@aldenpotamus I think that you press the reroll button a lot more than I do, and I think it's because of the advanced training omen. You have a much higher chance of getting a blue "thank god it's not reliability" than I do. So if I see a blue reliability, I'm not going to press reroll since I'm likely to see 3 non-rare options and then 2 more non-rare options.
@@abigailmcdowell4248 Do you run advanced training on almost every run? I think my feeling that reliability is good is tied to how I don't run that, so I find that pressing reroll to skip a blue reliability is a poor choice on my runs and end up taking it on many of my heroes. On runs where I do have it, I don't press reroll enough and that's why I stopped running it. I prefer not to rely on reroll but that's just a personal preference and fortunately the game accommodates with plenty of awesome omens to choose from.
@@TysonJensen I only started running the omen of better rerolls after watching alden talk about how good it was, but now it's one of my staple omens, especially on the later maps where you have a lot more omen slots and you're expecting to reach a much higher level, I 100% agree with you that the value of your rerolls matters massively, and that the first blue level up of reliability is hard to roll past, even more so without the better level ups. The way Alden has been evaluating them does lose that nuance, but just from the way the calculations work Reliability has the greatest diminishing returns, and doesn't work as synergystically with other stat boosts, it also always comes in the biggest increments especially on gear, so it's very easy to change your gear to get some other stats they might have and lose or gain 30-40% reliability, and any reliability over 100% is completely wasted, and so become completely wasted level ups (this is even more true if you have the perks that give you situational reliability, cherry picking and surgical strike)
Hey @AlekThunder47 - I hear you. To be honest I think I started this video out with a bit of a disclaimer that the details of the damage formula weren't a requirement for playing the game. That being said it's always a balance of trying create something complete and also brief. I think there's a quote that's often miss attributed to Mark Twain that goes "If I'd had more time I would have written you a shorter letter." or something that that effect. All that being said I understand where you're coming from, feedback received!
Glad to see this mamoth project finally be seen by the world! can't wait to see the whole project and your conclusions as it comes out!
Can't wait to hear what I got wrong! 🤣 but seriously thank you for all your help! I'm really pleased with how this first installment came out!
I saw a Last Spell guide and decided to click the video to leave a thumbs up just to support you because I have learned so, so much from your previous guides on the game.
I’m not math inclined and had no interest in a 42 minute video about numbers. Aside from skipping over some of the intro section, I watched the entire thing and was not bored for a moment. It didn’t feel like 42 minutes and was engaging the entire time. It was also explained in a way that was easily understandable.
I can’t wait for an expansion of some sort for this game. It is so fun and manages to scratch several gaming itches at once in a way no other can. Strategic turn based combat, rpg leveling and skills, arpg character loot, base and economy building, and tower defense all rolled into one uniquely engrossing game.
Thanks again for all the work you put into your videos. Listening to someone talk about a topic they’re extremely knowledgeable in is like chocolate for my brain, especially when it is something I’m interested in. I’ll once again say that the Last Spell subreddit should just be one post with a link to your RUclips channel.
Incredibly kind words and high praise Carl! Thank you! As for me I'll keep in mind that math based titles can definitely be a serious turn off for folks 😂 I had an amazing time making this video and am looking forward to getting back into the swing of things and making more before said expansion drops (can't wait)!
Again thank you and cheers!
the devs did announce DLC coming this year! haven't heard any more news about it for a while, but there is definitely more to come from this game
Appreciate the guide, love finding niche channels
Welcome on in! Glad to have you along for the fun!
I can’t even begin to imagine how much work and thought you and the great community you gathered put into this..
Great content, we love to see it
Thank you Zwerg! It was a fun project and I'm psyched that I managed to kick it out the door, now I just need to get back into a routine and pump out more content like this to counterbalance my new approach to the sub box which seems to be providing a minor haircut in views :)
Amazing!! Thank you to all the hard work that went into this. Quite the journey so far with the Last Spell, the game has come so far from alpha 🤩
It's a pretty spectacular game, perhaps a few bugs to iron out and more content would be quite amazing, but I've enjoyed every bit of the ride! Cheers!
I was running into some early damage problems in some of my harder runs, I'm really happy to finally see a tool to help solve this problem!
Definitely report back! Can't wait to hear if this actually move the needle for you @poimaster
Cheers!!
I do not usually comment, but this much work needs some algorithm boost)
Much appreciated, I'm glad you saw the content worth the comment. Feels good! Cheers!!!
I watched part 2 before this video. I will definitely test out the spreadsheet and give feedback.
Can't wait to hear what you think!
Awesome and informative delivery, your videos made my experience with the Last Spell a lot more enjoyable!
It's a wonderful game if sometimes a little hard to approach. It always means a lot to me to hear that my videos made folk's experiences that much more fun or rewarding!
Very good video, great to see that this beautiful game still has passionate people playing and mathing in it!
One thing I always observed is the one thing you said about the omen of dexterity: that it is basically the best omen,
I have always wondered why that is the case, whether I was at my early stages in playing and didn't like accuracy level ups or now when I take them mostly if it is blue
but chose damage/crit/ect upgrades if not blue and it hit me when I was watching the video, overwhelmingly strong when needed and useless when not was very well put
because the truth of this game is that the first 12 accuracy are overwhelmingly strong, so much so that 1 point is a steal for an omen that gives accuracy
In another way that I now think it, the accuracy omen allows you to spec in damage immediately
which is exactly what we want in this game as the enemies get tougher very quick on high apocalypse levels
The reason I bring accuracy up is because I had a realization about reliability which I would like to see yours and others take on it
I'll preface by saying that your rant about reliability was very funny and enlightening, please do more rants on stats in the future!
My main take, just like accuracy reliability is overwhelmingly strong when needed and useless when not
Now please do bring the pitchforks yet, there may be a method to my madness
Here is the thing, what do we need to do in this game: to kill enemies, and the best way to kill enemies is in one hit( ie one skill use )
And how do we kill enemies in one hit, by making sure that we roll damage in our damage range that is above the monster's effective hp( hp/armor/block/total ect )
Reliability does just that, increases the minimum damage just enough to kill enemies in one hit
Not related to reliability, but the reason why HC is so monstrous becuse it can kill 7 enemies in one AP use and even refund the AP it just used!
Average damage on the other hand, reliability's nemesis, is only valuable against bosses and perhaps elites in early nights
I've never found any other enemies that I need to do more than 3 skill uses
Big enemies usually die easily to momentum/multi-hit skills or one rifle shot,
although bosses usually also die to one rifle shot so I should probably not include that in the comparisons
If when we have reliability and we don't kill in one hit, reliability is basically useless, every other damage stat is almost always better
Also I should say that reliability is good only for 5 weapons because of damage ranges: 1h axe, 2h axe, the two bows and tome, it may be okay on 2h sword and wand
2h axe with the shortbow are probable the best with respect to reliability, level 5 has 121-236, damage range of 115 which is almost equal to its min damage of 121
i.e 10% reliability is almost the same as 10% damage for its min damage, that is very high to close in on its one hit potential
shortbow level 5 96-192: exact 96 damage range, same reasoning as the 2h axe applies
Since the damage ranges of weapons are the biggest in later rounds, reliability can be looked at as the late game accuracy for some weapons( accuracy of damage more specifically )
and probably the biggest reason reliability is so poor in comparison to other stats
late game you will have gear that just gives it on the side, and since reliability has the biggest diminishing returns
taking it as a level upgrade is less lucrative
And when in late game it is very likely that you no longer need the extra damage as you already have enough damage for one hits
Accuracy for instance always rolls in its full range 0 - 100 ( or 5 - 100 accounting the dodge rules ), does not matter how much you have of it
Reliability on the other hand does not follow that rule, or if someone has observed they can keen me in on how reliability rows are calculated
But in essence with 0 reliability you have 0 - 100% of damage range difference, in the case of the shortbow it would be from 0 to 96 extra damage
a 20% increase makes it from 20 - 100%, however now you are rolling 80 values instead of the full 100, i.e reliability is now 80% effective
translated in another way 20% reliability is 16% if this makes any sense
50% will have rolls from 50 - 100, now only 50 values to roll, no such problem for accuracy which is why it is good
If for instance 20% allowed rolls from 20 to 120 and everything above 100 converted into max damage now it would be much much better
imagine that now 50% reliability for instance allows max damage rolls 50% of the time, how good does it sound now?
All in all I really wish reliability was better then it is, however it seems to me that for a few weapons it is actually very good
and a good rule I have is to have somewhere around 20% before it starts having massive diminishing returns, usually look to have it from items
Omen of dexterity helps both in the early and late game. In the early game, you have a limited amount of power so missing a shot feels particularly awful. In the late game, every hero got to pick something other than accuracy on level-up at least once because you had omen of dexterity. Hard to quantify exactly how much power that is, but it's noticeable.
If you are not doing corpse piles, it doesn't help as much early game since you just rush the enemy and kill as fast as possible, but corpse piles are so good that not doing them is running the game on super-hard mode. When you have to just sit around waiting for the enemy to arrive you can't afford misses when the enemy finally enters the haven. (yes, I soften them up as I can, but I have AP remaining on each of the first few turns waiting for enemies to be where I want to actually kill them.)
So I very very rarely run omen of dexterity, especially early the only enemies that have any dodge that is worth worrying about are dodgers, and it's only on ranged damage weapons that accuracy is a significant issue, I think it's also more okay than you think to miss like 4 killing shots in a row and waste 4 AP, it happens to me a lot, and is very very annoying, but this game rarely requires perfection and even more rarely rewards you for it.
Reliability really does feel like a very dissapointing stat, the greatest value I can think of it providing is that it can increase the chance of you putting an enemy into a badly wounded state rather than just a wounded state without increasing your chance to kill, but the situations in which that's important are very very slim.
Reliability is also pretty abundent from perks as well as gear, I often take cherry picking at tier 1, an interesting wrinkle to that is that Longer weapons, a staple perk, gives you negative reliability, but reliability is capped at 0%, so if you never take reliability at level up or on gear, then Longer weapons has absolutely no downside!
The thing I brought up with alden when helping with this project, which is something you were very close to discussing here, is that the thing we actually care about isn't max or average damage, but %chance to kill, and I would be very interested to see how reliability fares in that, because things that increase your average/max damage also increase your overkill damage, but if less than max damage kills, than reliability has a greater comparative value, the issue with doing that is quantifying it would require simulating full night enemy compositions including their stat scaling and looking at what skills kill them in how many uses, which is so so so much more work that at the moment it's just not feasible
Hey @florfall ! So first off expect stat by stat rants coming very soon... and i apologize in advance for them because they will all likely be a little unhinged.
So I think the thing about reliability that really broke my heart more than anything was nearly exactly what you're describing. In my mind reliability *should* be the best stat for ensuring that something you *can one-shot* you will one-shot, and sadly it's not even good at that.
The reality is that you average specific damage upgrade is exactly as good as reliability at raising minimum damage as reliability... and this is when considering the absolute best case scenario, weapons with a very high damage range. Being able to put number to this phenomenon was the ah-hah moment that really soured me on Reliability entirely. If you can get more minimum, maximum and average damage from one stat and the other provides no maximum and less average damage while only providing the same minimum damage is basically a dead stat in my eyes.
To speak to the specific math that you've shared... I think I got lost a bit, which is not to say you don't have a point. The way I conceptualize Reliability is to simply imagine a graph with two lines one charting min damage and the other charting max damage, reliability simply creates a diagonal line between the two shaving the bottom of the region off as min damage converged on max damage. In practice the difference between min and max damage never really get's the extreme and the total distance that it's closing is relatively small.
But here's my idea... and this one is going to take some doing so it won't happen overnight. My plan is to write a modded game file that doesn't change the game at all. Instead it just records every hit and notes the ones that could have been impacted by reliability, aka when could a 1-shot have been possible with reliability but was missed. This would illuminate exactly how much reliability could actually be a difference maker. My hypothesis is that maybe around night 3-4 there is a spike where you aren't yet doing "excessive" damage and you'll get a slightly elevated rate of "could have killed" happening and then that too will die down. So we'll see when I make that and who will test it with me! :)
I generally agree Tyson, I think that Omen of Dex is the absolute goat because it really saves you two levels of accuracy to get to what I consider to be the "absolute minimum" you need to function in the mid game. For 1 omen point I'm getting the freedom to pick two level-ups that are tailored to the hero, because end of the say every hero need accuracy.
On the topic of accuracy @abigailmcdowell4248 ... my thinking is pretty simply -- how often will I not take an AP level-up. Answer: I will always take one. So conversely missing a hit is often like losing one AP. Every hero needs accuracy and 12% accuracy is really in the sweet spot to remove 95% of misses in the early game when AP is most at a premium. Having played with the spreadsheet, even enemies having 5% dodge causes accuracy to be the best stat by far... so I think that really has cemented in my mind that Omen of Dex is just king of the castle :) You should definitely try it, the opportunity cost is so low and the value is off the freaking charts :)
An interesting thing about the local maxima problem; if you know what level you are going to end up (for example buying a hero from the inn on the final day) it would be useful to find the true maxima for that given level, because there is no future you need to plan for
So I think my thinking still holds... if you consider a truly extreme example you could imagine a damage field where the hero is worthless until the final level and then BAM they become super human because the slope is so steep on that last level. Obviously with the current dynamics that's not possible... but the idea of ending up slightly stronger for a weaker ascent is still probably not worth it. We want to be winning at every level and rarely is the final level destination worth the sacrifice in the midgame.
I saw in a previous video you said something about the future for the game not looking good as far as updates. But I also read a supposed developer comment on reddit about working on a dlc. So hoping that is happening! Just discovered the game and am in love.
DLC was announced by the devs in the discord! it's just it's been quite a long time since there was a new patch, and there have been far fewer balance changes since releaase than a lot of other games in this space get
Hi @someguy8070 ! Like Abigail said the dev have promised DLC this year, so we're all eagerly anticipating more content and changes. The developers have been quite busy with a few other releases and for a while The Last Spell was feeling a little neglected (there are a few minor bugs here and that that I'd love to see patched), but I think the future of the game is bright assuming the devs deliver on their promises!
In other news we (the larger community on discord) have recently been breathing some new life into the game with modding that's been a lot of fun as well!
It's a good game to fall in love with, you're in good company :)
Thx for your effort now I know why sometimes I can oneshot some stuff and why not 😂 Like your vids they are very informativ😮
So happy to hear they help! I do think that exactly what decisions we make and how they contribute to damage can be a pretty mysterious relationship. Hope this helps and glad you're enjoying the vids!
And what about investing in poison or action points? How to calculate benefit?
@9237 so AP is almost always good with the diminishing return coming around 10AP... as for poison - it's a stand alone stat that doesn't scale in concert with anything else, so if you're using poison more is better and at 400% poison it's surprisingly effective without being quite as strong as some of the other options.
I feel like we have traumatised you with our jokes about reliability haha
There is no question... that cut to editor Alden was the product of more PTSD than any person should feel over such a worthless stat.
@@aldenpotamus And of course you always run the omen of "please, please can I have something other than reliability, please, please with sugar on top" 🤣
Omen of "Level-ups are Roughly 60% of Hero Power By the End of a Run So Why Not Take Good Ones" as I like to call it ;)
Thx for the video!
Thank for tuning is as always Alex!!! :)
Many much thanks.
@rosselliott6515 you are very welcome!
Blue reliability is 20% and is easily better than what is typically offered against it. green / beige reliability is generally worse than anything offered against it. Obviously if the blue reliability is up against say, +1 multi-hit, well. I usually *don't* run the "improved level-up odds" perk which makes reliability a lot better than it is if you have the improved odds on re-roll. Alden *always* runs improved level-up so reliability falls off pretty fast.
so 20% reliability is often not that much aditional damage? the change in the lowest possible damage can be fairly significant, but the change in average damage isn't that much, the spreadsheet and calculations that Alden is doing does take into account that you are getting 20% reliability vs 8% physical damage for example, and even with that reliability is never fantastic,
I do think that one of the things most lacking from aldens spreadsheet is some expected value of a reroll.
I do think that when you look at how the damage numbers shown in the spreadsheet change when you look at a reliabilty vs and opportunism or isolation level up it's pretty extreme, and then you need to consider that green level ups are closer to blue level ups than grey, so I would still broadly agree with Alden that a blue reliability is not great compared to the other stats that improve your damage, I do think there is something to be said for the improved consistency that reliability provides, but it's very hard to assign that a worth.
Then a thing that I am hoping to look into with aldens help soon is how much of each stat you can expect to be given from the random attributes that appears on stats, and reliability is one of the stats that gets the biggest numbers from gear, and so recieves even greater diminishing returns from level ups (the biggest example of that for me is the cape that gives 33% reliability at level 5 with no rarity, getting you 1/3 of the way to the cap in one item is huge and massively devalues getting that resource from other places)
Alden is very very aware of how contentious Reliability is, and I'm expecting it to get a lot of the attention in the upcoming videos that dive deeper into how the spreadsheet works
Hey Tyson! As always excited to see a comment from you, super appreciate you thoughts on these topics! I've been giving some thought to the hypothesis that advanced training (and my regular use of it) might be coloring my opinion on Reliability and I'm not so sure that it is. The analysis that I did generally considered the highest level of upgrade (simply because "more of the best thing" will always be better than "less of the same thing").
As I mentioned in the video precents are precents when it comes to the value of stat upgrades with the exception of reliability. Since that 20% is operating only on a portion of total damage (specifically the delta between min and max) it doesn't have nearly the magnitude of other stats like physical damage or opportunism where 1% is indeed (ignoring diminishing returns) 1% improvement of your overall damage.
I think the thing that was most damning about my reliability analysis was that a comparable rarity improvement of reliability or physical damage doesn't yield more min damage let alone more average damage. As a result it's nearly always better to choose something that will improve your min and max damage as you get twice the impact on average damage and seemingly also a proportional increase to min damage.
All this being said, I spent a lot of time in the sheet trying to find places were Reliability was the top dog and while they exist they're pretty rare when you consider a 12-ish level cap... so for now I'll happily take reliability on solid gear pieces but also likely not take it from level-ups except in the most extreme of cases.
As always super appreciate the comment! Cheers!
@@aldenpotamus I think that you press the reroll button a lot more than I do, and I think it's because of the advanced training omen. You have a much higher chance of getting a blue "thank god it's not reliability" than I do. So if I see a blue reliability, I'm not going to press reroll since I'm likely to see 3 non-rare options and then 2 more non-rare options.
@@abigailmcdowell4248 Do you run advanced training on almost every run? I think my feeling that reliability is good is tied to how I don't run that, so I find that pressing reroll to skip a blue reliability is a poor choice on my runs and end up taking it on many of my heroes. On runs where I do have it, I don't press reroll enough and that's why I stopped running it. I prefer not to rely on reroll but that's just a personal preference and fortunately the game accommodates with plenty of awesome omens to choose from.
@@TysonJensen I only started running the omen of better rerolls after watching alden talk about how good it was, but now it's one of my staple omens, especially on the later maps where you have a lot more omen slots and you're expecting to reach a much higher level, I 100% agree with you that the value of your rerolls matters massively, and that the first blue level up of reliability is hard to roll past, even more so without the better level ups. The way Alden has been evaluating them does lose that nuance, but just from the way the calculations work Reliability has the greatest diminishing returns, and doesn't work as synergystically with other stat boosts, it also always comes in the biggest increments especially on gear, so it's very easy to change your gear to get some other stats they might have and lose or gain 30-40% reliability, and any reliability over 100% is completely wasted, and so become completely wasted level ups (this is even more true if you have the perks that give you situational reliability, cherry picking and surgical strike)
So much ranting. Could've put everything into 5 to 7 minutes.
Hey @AlekThunder47 - I hear you. To be honest I think I started this video out with a bit of a disclaimer that the details of the damage formula weren't a requirement for playing the game. That being said it's always a balance of trying create something complete and also brief. I think there's a quote that's often miss attributed to Mark Twain that goes "If I'd had more time I would have written you a shorter letter." or something that that effect.
All that being said I understand where you're coming from, feedback received!