Spreadsheet link for Members (members, leave a comment on this video letting me know you requested access to view document and I will grant it ASAP): docs.google.com/spreadsheets/d/17Io_Xb8QrDljR2RWOyD3_16H3sZKvg8DcWvTD-z5muU/edit?usp=sharing Hit that join button if you want access!
I was kinda shocked to see Raisel Iglesias so low on the RPC scale, considering he went on that no baserunner tear, but like Clase he isn't a high strikeout pitcher and you said the scale ranks high strikeout guys higher
I love that baseball finally embraced statistics. I was brought up (and played) in the “you can’t quantify what baseball players do” time; and it was willful stupidity to think like that.
Man I'm only 23 but when I was a kid I remember Brian Kenny arguing with others on MLB Network that pitcher wins and losses were bad stats. Mad Dog and the old guard loved to argue that good pitchers will inspire their lineups to score more runs. Most ridiculous thing I ever heard.
Clase’s postseason struggles show why ERA and saves don’t tell the whole story. It’s time we start looking deeper at what really makes a reliever elite.
As a Yankees fan who grew up watching and idolizing Mariano, it's not super wild in the proper context. A lot of guys have been better than Mo. Lots of young studs come into the league and have a few seasons better than Mo. A lot of guys have put things together and peaked higher than Mo ever did. The difference is that Mo did it for almost 20 years. All of those other guys were better for a time and then flamed out. He was a rock; consistently elite year after year at a notoriously volatile position. Clase is the latest in a long list of guys who were better than Mo for the start of their career as relievers, and everyone so far has fizzled out. That list includes the likes of Gagne, Papelbon, Britton, Kimbrel. Clase hasn't seen that fall off yet, so it's possible that he bucks the trend and ends up with a career better than Mo's. He's better than Mo right now, but we'll see in 15 years if he can keep it up.
Great concept for better understanding relievers!! I can't wait to see how this looks throughout the 25' season! Hope you make something like this for the hitters or maybe defense as well!!
A reliever statistic I thought of. The number of times a reliever did all three of the following: A: pitched at least one inning B: Allowed no runs And C: Allowed no more than one baserunner. I call that the definition of quality relief appearance. They have quality starts for starters, so how about this for relievers? This also could be expressed as a percentage of appearances where the reliever faced three batters perhaps.
They have holds a hold means when you came in your team haf a lead and when you left you still had a lead I think allowing 0 baserunners is a really high bar for a "quality appearance" stat tho
@@backyardbaseball2006I was thinking no runs allowed, and up to one baserunner allowed, and my line of logic is that it would probably weed out a good number of relief appearances where the pitcher pitched his way into trouble even if he ultimately got the save or hold.
Hello, And That's Baseball. I was wondering if anytime soon, you can do a video about the New York Mets of the 1960's. Unfortunately, Baseball programs on you tube don't discuss those teams. Growing up with those teams, I have wanted to see some exciting coverage from one of the youngsters on this app, but haven't had any luck finding a program discussing those teams. Thanks for taking the time to read this! Signed, Hank.
I feel like that topic doesn't really fit my format as much since I'm trying to lean more into analysis, but I'll share the idea with some of my friends who do more historical content. The 1962 Mets are a remarkable story on their own, and I'd also be glad to see someone cover their progression from that to winning in '69.
Fun video. Glad to see you bringing up May's video. Seeing someone trying to find consistency in the ever inconsistency in relievers is always fun and I hope your stat shows its truth in the upcoming season. Also, I am glad to see Clay Holmes be given a pass as the more I look at his stats, the more it seemed that he got unlucky one way or another. One thing I am curious about is whether the stats would correlate with knuckleball pitchers as your stat relies on stuff that doesn't place high value on the knuckleball(Although there was only Waldron who is a starter).
I feel like knuckleballers are so anomalous that they need to be examined independently regardless. I really don't know how this formula would treat them. Could test it on starters in the future and see where Waldron grades out.
If rpc became the norm for evaluating relife pitchers it would be interesting if this effected relief pither contracts and the amount of relief pitchers in the hof
After watching the whole video, I think your RPC methodology was pretty strong and including the failsafe for potential shortcomings in stuff+ models was a great idea. To refine the weights in the formula you could analyze p-values to get a good idea of the predictive value of each component of RPC. Regardless this is great stuff, nice video!
With the caveat that I’m only 3 minutes into the video, I think it’s a bit misleading to say ERA doesn’t matter. For relievers it doesn’t hold much predictive power I 100% agree, but the primary purpose of any predictor, whether results-based or not, is that it is meant to predict future performance in run prevention. ERA is still imperfect for relievers, when considering inherited runners, win probability added or lost, etc, but ERA is still a good barometer for reliever value in any given year, even if it isn’t very sticky or an indicator of real talent
I always thought that inherited runs that score should be split between the two pitchers. Punishes the first pitcher less and gives a better indication of how effective a reliever really is.
While I agree overall saves in the modern era aren't important, I ironically and maybe contradictive to myself disagree about save rate not mattering. Its the most surface level high pressure pitching metric out there, but it still catches an eye when you see a guy who had like 8/9 saves locked down and makes you wanna see the more niche "how does he do in the 7th with runners on and 2 outs" or whatever
I think there's a discussion to be had about leverage/role mattering, but I haven't seen any evidence to prove it so far. I also think as we move more towards fluid bullpen usage that fewer elite relievers will be saved for purely 9th inning roles. A guy like Clase would've probably entered a lot more high leverage situations in the 8th or even 7th this year if the rest of their bullpen wasn't so elite.
Before I sit through this whole thing there's a BIG. SHINY. RED FLAG already. There's this... Thing we have in baseball that has come about in the past decade. We have a bunch of people coming into baseball fandom who never played the game seriously, don't really have any interest in swinging a bat, or the nuances of where you should step at shortstop when the pitcher is delivering the pitch. They are solely interested in the mathematical side of baseball. To be clear, that's awesome, it's good for the game, it's good for everyone. Enjoy. But sometimes these people say "results based stats" like it's a swear word. Let's be really, really clear. Baseball is a sport. It's a competition. At the end of the day the only thing that actually matters is who scored more runs. Who won the game. Everything else, from home run records, to hitting for the cycle, to a perfect game, it's all secondary to that. It's all in persuit of winning. If you ask any guy in MLB if he would rather go 4 for 4 and lose, or go 0 for 4 and win... Well they would probably tell you they want to go 4 for 4 and win, but if that's not an option they will take the win. I say this to explain why results based stats are THE ONLY stats that matter when you are looking back and judging a player's performance. Advanced analytics are great for front offices trying to predict which players are going to be good in the future. But at the end of the day if a guy has an xFIP of 10.00 for his entire career, but he posts a 2.00 career ERA he was great. If he has an xFIP of 2.00 and he posts a career ERA of 10.00 he had a terrible career. Why? Because "expected" stats don't do any good whatsoever towards the goal of winning the game. In the case of a pitcher you win by pitching as many innings as possible and allowing as few runs as possible. If you were trying to convince me why I shouldn't sign a guy NEXT season because his expected numbers are considerably worse than his actual results, fine. Good point. But if you are trying to tell me a guy was bad LAST YEAR because his expected stats were worse than his results that's totally irrelevant. He got the results. It doesn't matter if he got lucky, or if he used a combination of actions that are not likely to be reliably successful in the future. That's all completely meaningless. Those games are already in the books. They are already won or lost. So all that matters now is the RESULTS.
Spreadsheet link for Members (members, leave a comment on this video letting me know you requested access to view document and I will grant it ASAP): docs.google.com/spreadsheets/d/17Io_Xb8QrDljR2RWOyD3_16H3sZKvg8DcWvTD-z5muU/edit?usp=sharing
Hit that join button if you want access!
This is real nerd shit, I absolutely love it. I think RPC works very well.
We’ll see how that ranking looks this time next year
Why I love baseball
I was kinda shocked to see Raisel Iglesias so low on the RPC scale, considering he went on that no baserunner tear, but like Clase he isn't a high strikeout pitcher and you said the scale ranks high strikeout guys higher
He's still top 20, which is very high considering how strong of a relief season 2024 was
I love that baseball finally embraced statistics. I was brought up (and played) in the “you can’t quantify what baseball players do” time; and it was willful stupidity to think like that.
Man I'm only 23 but when I was a kid I remember Brian Kenny arguing with others on MLB Network that pitcher wins and losses were bad stats. Mad Dog and the old guard loved to argue that good pitchers will inspire their lineups to score more runs. Most ridiculous thing I ever heard.
Clase’s postseason struggles show why ERA and saves don’t tell the whole story. It’s time we start looking deeper at what really makes a reliever elite.
To clarify, I still think he’s elite. It’s more about ERA being incredibly high variance. He wasn’t so good that he was completely untouchable.
“He’s better than Mariano” is the wildest thing a MLB player can say. Come on Hegdes, you should know better
It's out of context tbf but he still didn't have to make that comp at all lmao
As a Yankees fan who grew up watching and idolizing Mariano, it's not super wild in the proper context. A lot of guys have been better than Mo. Lots of young studs come into the league and have a few seasons better than Mo. A lot of guys have put things together and peaked higher than Mo ever did. The difference is that Mo did it for almost 20 years. All of those other guys were better for a time and then flamed out. He was a rock; consistently elite year after year at a notoriously volatile position. Clase is the latest in a long list of guys who were better than Mo for the start of their career as relievers, and everyone so far has fizzled out. That list includes the likes of Gagne, Papelbon, Britton, Kimbrel. Clase hasn't seen that fall off yet, so it's possible that he bucks the trend and ends up with a career better than Mo's. He's better than Mo right now, but we'll see in 15 years if he can keep it up.
One of my favorite stats for pitchers is WHIP. I would be curious to see the R-squared value for it in your calculations.
We heard that “Better than Mariano” shit and shut that shit down real fast in the playoffs lol.
Great concept for better understanding relievers!! I can't wait to see how this looks throughout the 25' season! Hope you make something like this for the hitters or maybe defense as well!!
I think wOBA/wRC+ are good enough as one number stats for hitters, even if they're influenced somewhat by luck
Another incredible video. Thanks for making these!
Glad you liked it!
A reliever statistic I thought of.
The number of times a reliever did all three of the following:
A: pitched at least one inning
B: Allowed no runs
And C: Allowed no more than one baserunner.
I call that the definition of quality relief appearance. They have quality starts for starters, so how about this for relievers?
This also could be expressed as a percentage of appearances where the reliever faced three batters perhaps.
They have holds a hold means when you came in your team haf a lead and when you left you still had a lead I think allowing 0 baserunners is a really high bar for a "quality appearance" stat tho
@@backyardbaseball2006I was thinking no runs allowed, and up to one baserunner allowed, and my line of logic is that it would probably weed out a good number of relief appearances where the pitcher pitched his way into trouble even if he ultimately got the save or hold.
Hello, And That's Baseball. I was wondering if anytime soon, you can do a video about the New York Mets of the 1960's. Unfortunately, Baseball programs on you tube don't discuss those teams. Growing up with those teams, I have wanted to see some exciting coverage from one of the youngsters on this app, but haven't had any luck finding a program discussing those teams. Thanks for taking the time to read this!
Signed, Hank.
I feel like that topic doesn't really fit my format as much since I'm trying to lean more into analysis, but I'll share the idea with some of my friends who do more historical content. The 1962 Mets are a remarkable story on their own, and I'd also be glad to see someone cover their progression from that to winning in '69.
WHIP is the first and sometimes only stat I look at when evaluating relievers.
Wholeheartedly agree. WHIP is the most important stat of a Pitcher. How many hits and walks do you give up?
We love quantitative analysis r-squared tests 😤
RAHHHHHH 🦅
Fun video. Glad to see you bringing up May's video. Seeing someone trying to find consistency in the ever inconsistency in relievers is always fun and I hope your stat shows its truth in the upcoming season. Also, I am glad to see Clay Holmes be given a pass as the more I look at his stats, the more it seemed that he got unlucky one way or another. One thing I am curious about is whether the stats would correlate with knuckleball pitchers as your stat relies on stuff that doesn't place high value on the knuckleball(Although there was only Waldron who is a starter).
I feel like knuckleballers are so anomalous that they need to be examined independently regardless. I really don't know how this formula would treat them. Could test it on starters in the future and see where Waldron grades out.
@@AndThatsBaseball Interesting and most likely true. Great work.
2013 Koji Uehara would be a very interesting analysis
Incredibly good pitcher who also had luck go his way that year. Fantastic season all around.
If rpc became the norm for evaluating relife pitchers it would be interesting if this effected relief pither contracts and the amount of relief pitchers in the hof
3:49 absolutely love that little bit of Jacob deGrom propaganda thrown in there.
its a good day when atb uploads a fire video
Thanks goat
He had the KENLEY JANSEN EFFECT
After watching the whole video, I think your RPC methodology was pretty strong and including the failsafe for potential shortcomings in stuff+ models was a great idea. To refine the weights in the formula you could analyze p-values to get a good idea of the predictive value of each component of RPC. Regardless this is great stuff, nice video!
I’m glad you enjoyed. Will definitely take this into consideration while tweaking this formula in the future.
lol I made it here within 30 seconds of uploading
With the caveat that I’m only 3 minutes into the video, I think it’s a bit misleading to say ERA doesn’t matter. For relievers it doesn’t hold much predictive power I 100% agree, but the primary purpose of any predictor, whether results-based or not, is that it is meant to predict future performance in run prevention. ERA is still imperfect for relievers, when considering inherited runners, win probability added or lost, etc, but ERA is still a good barometer for reliever value in any given year, even if it isn’t very sticky or an indicator of real talent
I always thought that inherited runs that score should be split between the two pitchers. Punishes the first pitcher less and gives a better indication of how effective a reliever really is.
While I agree overall saves in the modern era aren't important, I ironically and maybe contradictive to myself disagree about save rate not mattering.
Its the most surface level high pressure pitching metric out there, but it still catches an eye when you see a guy who had like 8/9 saves locked down and makes you wanna see the more niche "how does he do in the 7th with runners on and 2 outs" or whatever
I think there's a discussion to be had about leverage/role mattering, but I haven't seen any evidence to prove it so far. I also think as we move more towards fluid bullpen usage that fewer elite relievers will be saved for purely 9th inning roles. A guy like Clase would've probably entered a lot more high leverage situations in the 8th or even 7th this year if the rest of their bullpen wasn't so elite.
To me the foundation of a good reliever is FIP, Whip, IP and K/9
Regular season, & playoffs, are different stat categories.
RYAN WALKER MENTIONED
He's really good
I hate having to re watch all of this, but i know you know ball and want to know what you think
Grace period has expired, time for analysis
Socks to bed=best sleep possible 🟥🟧🟧🟪🟩🟥🟨🟨
Simple, he can’t handle pressure. Some people thrive under pressure, some people Kershaw under pressure
Clase was pretty bad in 23.. no getting around that
222
too much math blah yawn
Before I sit through this whole thing there's a BIG. SHINY. RED FLAG already. There's this... Thing we have in baseball that has come about in the past decade. We have a bunch of people coming into baseball fandom who never played the game seriously, don't really have any interest in swinging a bat, or the nuances of where you should step at shortstop when the pitcher is delivering the pitch. They are solely interested in the mathematical side of baseball. To be clear, that's awesome, it's good for the game, it's good for everyone. Enjoy. But sometimes these people say "results based stats" like it's a swear word. Let's be really, really clear. Baseball is a sport. It's a competition. At the end of the day the only thing that actually matters is who scored more runs. Who won the game. Everything else, from home run records, to hitting for the cycle, to a perfect game, it's all secondary to that. It's all in persuit of winning. If you ask any guy in MLB if he would rather go 4 for 4 and lose, or go 0 for 4 and win... Well they would probably tell you they want to go 4 for 4 and win, but if that's not an option they will take the win.
I say this to explain why results based stats are THE ONLY stats that matter when you are looking back and judging a player's performance. Advanced analytics are great for front offices trying to predict which players are going to be good in the future. But at the end of the day if a guy has an xFIP of 10.00 for his entire career, but he posts a 2.00 career ERA he was great. If he has an xFIP of 2.00 and he posts a career ERA of 10.00 he had a terrible career. Why? Because "expected" stats don't do any good whatsoever towards the goal of winning the game. In the case of a pitcher you win by pitching as many innings as possible and allowing as few runs as possible. If you were trying to convince me why I shouldn't sign a guy NEXT season because his expected numbers are considerably worse than his actual results, fine. Good point. But if you are trying to tell me a guy was bad LAST YEAR because his expected stats were worse than his results that's totally irrelevant. He got the results. It doesn't matter if he got lucky, or if he used a combination of actions that are not likely to be reliably successful in the future. That's all completely meaningless. Those games are already in the books. They are already won or lost. So all that matters now is the RESULTS.