Pressure Situations

sreem · 05-26-2015, 01:00 PM

I was wondering if anyone knows if there is a hidden rating for all players factors into how they perform in pressure situations. I.E. - runners on 2nd and 3rd, down by 1 for a batter. Runner on 3rd representing tying run for a pitcher.

endgame · 05-26-2015, 01:18 PM

I watch the stats for particular players who appear to perform better in pressure situations, i.e. 'clutch' hitters, pitchers with poise. The lines like runner in scoring position, bases loaded, first pitch, etc... actually there are many that address all situations I tend to look for, but those are just stats, indicators, not predictors. Regardless, I enjoy watching them compare to performances in key situations.

That being said, stats don't drive performance, they're just records of it. And if I could find it, then it wouldn't be hidden.

My guess is not, other than perhaps a tweak in key playoff games. FWIW

SteveP · 05-26-2015, 01:36 PM

The only thing I've ever seen, which might be an example, is when there is a runner on 3B with less than two outs, in the bottom of the ninth, or later inning, where that runner would score a walk-off run. The infield would be playing in. Normally, with a ground ball hit to one of the infielders, the runner would not advance in that situation, ever. But in the situation I've described, the runner may try to advance and score the winning run.

Other than that, I have to reason to believe that the game or AI works differently in a pressure situation.

sreem · 05-26-2015, 01:41 PM

Quote:

Originally Posted by SteveP

The only thing I've ever seen, which might be an example, is when there is a runner on 3B with less than two outs, in the bottom of the ninth, or later inning, where that runner would score a walk-off run. The infield would be playing in. Normally, with a ground ball hit to one of the infielders, the runner would not advance in that situation, ever. But in the situation I've described, the runner may try to advance and score the winning run.

Other than that, I have to reason to believe that the game or AI works differently in a pressure situation.

Note sure if the coding is possible, or if what I would call "poise" is even quantifiable. It seems there are some players you would rather have up or on the mound in key situations than others, but it could just be based on past performance instead of any quantifiable measurement of consistency "under pressure".

RchW · 05-26-2015, 01:41 PM

This is old data but I've yet to see anything recently that confirms clutch other than relying on your best players. This particular study compared fan picks of their best clutch hitters vs Tom Tango picking (some tongue in cheek) the teams best player. Very interesting.

With the game on the line, I want ? ? The Hardball Times

Edit. Clutch hitting exists but because the sample sizes are by definition small it is a better strategy to go with the better player (ie more stable performance) when you have the choice. There is data that shows power hitters fail in the clutch more often than singles hitters.

Edit 2:

Another Fangraph article from 2014 which gives a useful summary of previous work.

endgame · 05-26-2015, 04:01 PM

Quote:

Originally Posted by sreem

Note sure if the coding is possible, or if what I would call "poise" is even quantifiable. It seems there are some players you would rather have up or on the mound in key situations than others, but it could just be based on past performance instead of any quantifiable measurement of consistency "under pressure".

Oh yeah, couldn't agree more, but it's a managerial judgment. THE number one reason I lobbied for so long for the return of the IRS% to help me evaluate my RP confidence.

rrrozz · 05-26-2015, 04:07 PM

Quote:

Originally Posted by sreem

I was wondering if anyone knows if there is a hidden rating for all players factors into how they perform in pressure situations. I.E. - runners on 2nd and 3rd, down by 1 for a batter. Runner on 3rd representing tying run for a pitcher.

For pitchers I would think that "Hold" would come into play here.

NoOne · 05-26-2015, 05:58 PM

if there is empirical evidence it exists in RL, then it's possibly in the game.

the problem is you'd likely have too small sample size to know for sure. you'd have to go across too many years on a particular player to get enough data. then, you'd have to figure a way to weight data from year to year as the player changes. which, imo, involves subjectivity and taints the results.

Clutch with RiSP example:
they say an average normalizes around 1000AB (a recent thread linked a page with this info, not made up, it was stated as 991AB actually).

how many years before you have 1000 ab with RiSP? how much changes over that time that can influence results (league talent, aging player, league trends, rule changes etc etc)? ceterus paribus does not apply, so it's a complicated thing to discern.

the more unusual and rare the situation, the more difficult to determine if someone is 'clutch' or just had good timing for a hit based on his overall ability.

imo, in real life i think some players do let psychological aspects of the game affect them more than others, and it definitely affects results. i think it would be difficult to represent this in a game in an accurate way. there are fast and messy ways to do it, but the right way would be extremely complicated and likely a bit subjective in nature.

05-26-2015, 01:00 PM	#1
sreem All Star Reserve Join Date: Jul 2013 Posts: 673	Pressure Situations I was wondering if anyone knows if there is a hidden rating for all players factors into how they perform in pressure situations. I.E. - runners on 2nd and 3rd, down by 1 for a batter. Runner on 3rd representing tying run for a pitcher. __________________ "What doesn't kill you only makes you stronger. Except for bears....."

05-26-2015, 01:18 PM	#2
endgame Hall Of Famer Join Date: Aug 2002 Posts: 16,842	I watch the stats for particular players who appear to perform better in pressure situations, i.e. 'clutch' hitters, pitchers with poise. The lines like runner in scoring position, bases loaded, first pitch, etc... actually there are many that address all situations I tend to look for, but those are just stats, indicators, not predictors. Regardless, I enjoy watching them compare to performances in key situations. That being said, stats don't drive performance, they're just records of it. And if I could find it, then it wouldn't be hidden. My guess is not, other than perhaps a tweak in key playoff games. FWIW __________________ "Try again. Fail again. Fail better." -- Samuel Beckett _____________________________________________

05-26-2015, 01:41 PM	#5
RchW Hall Of Famer Join Date: Jul 2004 Location: Toronto ON by way of Glasgow UK Posts: 15,629	This is old data but I've yet to see anything recently that confirms clutch other than relying on your best players. This particular study compared fan picks of their best clutch hitters vs Tom Tango picking (some tongue in cheek) the teams best player. Very interesting. With the game on the line, I want ? ? The Hardball Times Edit. Clutch hitting exists but because the sample sizes are by definition small it is a better strategy to go with the better player (ie more stable performance) when you have the choice. There is data that shows power hitters fail in the clutch more often than singles hitters. Edit 2: Another Fangraph article from 2014 which gives a useful summary of previous work. __________________ Cheers RichW If you’re looking for a good cause to donate money to please consider a Donation to Parkinson’s Canada. It may help me have a better future and if not me, someone else. Thanks. “Conservatism consists of exactly one proposition …There must be in-groups whom the law protects but does not bind, alongside out-groups whom the law binds but does not protect.” Frank Wilhoit Last edited by RchW; 05-26-2015 at 01:55 PM.

05-26-2015, 05:58 PM	#8
NoOne Banned Join Date: Apr 2015 Posts: 7,273 Infractions: 0/1 (3)	if there is empirical evidence it exists in RL, then it's possibly in the game. the problem is you'd likely have too small sample size to know for sure. you'd have to go across too many years on a particular player to get enough data. then, you'd have to figure a way to weight data from year to year as the player changes. which, imo, involves subjectivity and taints the results. Clutch with RiSP example: they say an average normalizes around 1000AB (a recent thread linked a page with this info, not made up, it was stated as 991AB actually). how many years before you have 1000 ab with RiSP? how much changes over that time that can influence results (league talent, aging player, league trends, rule changes etc etc)? ceterus paribus does not apply, so it's a complicated thing to discern. the more unusual and rare the situation, the more difficult to determine if someone is 'clutch' or just had good timing for a hit based on his overall ability. imo, in real life i think some players do let psychological aspects of the game affect them more than others, and it definitely affects results. i think it would be difficult to represent this in a game in an accurate way. there are fast and messy ways to do it, but the right way would be extremely complicated and likely a bit subjective in nature. Last edited by NoOne; 05-26-2015 at 05:59 PM.

05-26-2015, 01:36 PM	#3
SteveP Moderator Join Date: Aug 2008 Posts: 3,109	The only thing I've ever seen, which might be an example, is when there is a runner on 3B with less than two outs, in the bottom of the ninth, or later inning, where that runner would score a walk-off run. The infield would be playing in. Normally, with a ground ball hit to one of the infielders, the runner would not advance in that situation, ever. But in the situation I've described, the runner may try to advance and score the winning run. Other than that, I have to reason to believe that the game or AI works differently in a pressure situation.