Production drop-off in playoffs

[Brad K]

Originally Posted by jcard

I would add that better documentation and description of many of the game mechanics would alleviate a lot of the confusion. For example, I have never seen an authoritative presentation of how TCR works. All players? Only players still developing? “Talent” is used interchangeably with “potential” on some screens / reports; is that also true for its usage in TCR?

Many features weren't documented when they were created and they've now forgotten what they did.

[krownroyal83]

I didn’t mean for this to turn into an argument. I basically wanted to know if there is a clutch factor in the playoffs. The smart, long term players of the game chimed in and have answered my question. There is no clutch factor. That’s all I needed to know. For immersion sake I’m gonna pretend there is. Makes for good story lines.

[Brad K]

Originally Posted by krownroyal83

I didn’t mean for this to turn into an argument. I basically wanted to know if there is a clutch factor in the playoffs. The smart, long term players of the game chimed in and have answered my question. There is no clutch factor. That’s all I needed to know. For immersion sake I’m gonna pretend there is. Makes for good story lines.

Yes, it does make for good stories. I write my league articles ignoring modern understanding and sabermetrics.

[PSUColonel]

Originally Posted by jcard

Does everyone feel better about themselves, now that they condescendingly regurgitated sabermetric orthodoxy in response to what was an honest question?

Moreover, the question was about OOTP, not real-life baseball. It is hardly heretical to ask the OP’s question regarding a sport simulation game. For example, in FM, FHM, and EHM there are mental ratings that impact the ability to perform in big games, etc.

I have seen no documented information or evidence to suggest that there is a type of clutch factor in OOTP. At the same time, while a reasonable starting point, blithely assuming that the OOTP engine always models in practice accepted analytical thought is a fraught default attitude.

Generally speaking, the deliberate opacity in the descriptions of several aspects of the game are sufficient to in many cases preclude the averment of beliefs as facts so regularly undertaken on this forum.

Correct. There is NOTHING in OOTP suggesting any reason for players to excel or not any more or less in the postseasons n compared to any other time of the year.

I think there are people who unfortunately engage in wishful thinking and/or fantasy when it comes to how OOTP operates. In fact I find it hilarious that people treat predictive stats in OOTP (such as FIP for example) the way they would in real life. It’s a complete waste of time since all stat output in the game is simply determined by ratings. It’s really no more or less complicated than that. Yes perhaps players are going against other players with better ratings in the postseason, but even then, certain things have to happen because of modifiers. So I’m a way statistical output is pre-determined as long as there aren’t talent changes. This doesn’t mean players won’t experience ups and downs, but over the long haul with ample sample size, the stats output will be a direct reflection of ratings. (Again..provided they don’t change)

I know some people around here think they are taking some kind of analytical deep dive when they sift through all of the ridiculous numbers when the only numbers that really matter are the ratings…pure and simple. It’s also not too difficult to know those numbers even with scouting on, so those are the only numbers you need to ever really focus on.


The way the OOTP engine works I believe mostly to be a case of Occam’s Razor.

[OutS|der]

Originally Posted by PSUColonel

Correct. There is NOTHING in OOTP suggesting any reason for players to excel or not any more or less in the postseasons n compared to any other time of the year.

I think there are people who unfortunately engage in wishful thinking and/or fantasy when it comes to how OOTP operates. In fact I find it hilarious that people treat predictive stats in OOTP (such as FIP for example) the way they would in real life. It’s a complete waste of time since all stat output in the game is simply determined by ratings. It’s really no more or less complicated than that. Yes perhaps players are going against other players with better ratings in the postseason, but even then, certain things have to happen because of modifiers. So I’m a way statistical output is pre-determined as long as there aren’t talent changes. This doesn’t mean players won’t experience ups and downs, but over the long haul with ample sample size, the stats output will be a direct reflection of ratings. (Again..provided they don’t change)

I know some people around here think they are taking some kind of analytical deep dive when they sift through all of the ridiculous numbers when the only numbers that really matter are the ratings…pure and simple. It’s also not too difficult to know those numbers even with scouting on, so those are the only numbers you need to ever really focus on.


The way the OOTP engine works I believe mostly to be a case of Occam’s Razor.

League totals Matter more than ratings. The game is driven by league totals with the help of the modifiers to adjust based on ratings.

IE drop a player with 250 power rating onto a league with 2023 HR totals and they will hit differently then if there were dropped into a league with deadball HR league totals.
Messing with league taken levels has less effect than messing with totals and the modifiers

[MathBandit]

Originally Posted by OutS|der

League totals Matter more than ratings. The game is driven by league totals with the help of the modifiers to adjust based on ratings.

IE drop a player with 250 power rating onto a league with 2023 HR totals and they will hit differently then if there were dropped into a league with deadball HR league totals.
Messing with league taken levels has less effect than messing with totals and the modifiers

This is part of the reason that I would disagree heavily with the idea that ratings matter more than something like FIP for a pitcher. Partly because of the context of LTMs, and partly because the ratings are from an inaccurate and biased source.

If I want to know if a 6-sided die is balanced, I'm not going to ask a high school maths student to measure each of the sides and provide me with a mathematical analysis of the balance of the die if I can just roll it 500 times instead and see the results.

[Sweed]

Originally Posted by PSUColonel

Correct. There is NOTHING in OOTP suggesting any reason for players to excel or not any more or less in the postseasons n compared to any other time of the year.

I think there are people who unfortunately engage in wishful thinking and/or fantasy when it comes to how OOTP operates. In fact I find it hilarious that people treat predictive stats in OOTP (such as FIP for example) the way they would in real life. It’s a complete waste of time since all stat output in the game is simply determined by ratings. It’s really no more or less complicated than that. Yes perhaps players are going against other players with better ratings in the postseason, but even then, certain things have to happen because of modifiers. So I’m a way statistical output is pre-determined as long as there aren’t talent changes. This doesn’t mean players won’t experience ups and downs, but over the long haul with ample sample size, the stats output will be a direct reflection of ratings. (Again..provided they don’t change)

I know some people around here think they are taking some kind of analytical deep dive when they sift through all of the ridiculous numbers when the only numbers that really matter are the ratings…pure and simple. It’s also not too difficult to know those numbers even with scouting on, so those are the only numbers you need to ever really focus on.


The way the OOTP engine works I believe mostly to be a case of Occam’s Razor.

I don't deep dive into saber stats in OOTP and I wouldn't argue they can be used exactly as they are in real life. Afterall OOTP is a computer game and real life is real life. Having said that I don't think one can totally dismiss their predictive value the way you have as a part of OOTP.

I'll stick with FIP since it was your example....

Isn't a player with higher ratings(skills) more likely to produce better stats?
Aren't those raw stats run through formulas to show us the "new" predictive saber stats? How do they not correlate? Shouldn't a P with very high ratings be more likely to produce a better FIP by the way the ratings interact and chance? Meaning high ratings are predictive of stat output? The stat output that gives us FIP?


Compared to real life OOTP is smoke and mirrors, ie code designed to give the illusion of a real baseball game being played. There are no physics involved but I don't think that means the smoke and mirrors numbers actually mean nothing. They still interact with the ratings and intricacies of the engine to produce numbers that have meaning. Otherwise what is the point? You yourself say "stats output will be a direct reflection of ratings. (Again..provided they don’t change)". And how is that different from real life? A P could be a league leader in FIP and in the prime of his career. If his skills(ratings in OOTP) haven't diminished then one can predict his performance. If his skills (ratings in OOTP) drop (velocity, movement, control) then his FIP will become less useful as a predictor. I see that as common sense whether IRL or OOTP. A pitcher loses 2 or 3 mph on his fastball and it starts to get hit harder in real life and his FIP goes up. In OOTP dice rolls are less likely to go that pitcher's way and his FIP goes up. Smoke and mirrors at work.

Seems to me in both real life and OOTP one can look at skill sets along with stat output to predict performance. In real life as well as OOTP these predictions are based on skills not changing. You actually say this yourself in regard to OOTP, "(Again..provided they don’t change).

[Sweed]

Originally Posted by Charlie Hough

Actually, no, everyone did not just condescendingly regurgitate sabermetric orthodoxy, and they did not blithely assume anything. The first few replies attempted to explain why the OP might be seeing the results described in OOTP, and those explanations are based on the actual conditions in question and how the sim works under those conditions. This comes from decades of combined experience with OOTP and a very thorough understanding of how the game engine works.

Originally Posted by jcard

1. You and I apparently read those replies differently.

.

Originally Posted by jcard

I regret using the word “everyone” in my original comment. I thought it would be understood that I was referring to the replies where a certain amount of smarminess.

And when reading your first reply I did take it as not actually meaning "everyone". Note that I made no reply to your first post.

I did take your second reply (read those replies differently) as doubling down on "everyone" and that is when I replied. My bad in misinterpreting you words/intentions there.

Thank you for clarifying.

[KRSW]

Originally Posted by Sweed

Isn't a player with higher ratings(skills) more likely to produce better stats?
Aren't those raw stats run through formulas to show us the "new" predictive saber stats? How do they not correlate? Shouldn't a P with very high ratings be more likely to produce a better FIP by the way the ratings interact and chance? Meaning high ratings are predictive of stat output? The stat output that gives us FIP?

Here is a video I recommend to you.
https://youtu.be/b2qimyrNK7M

[MathBandit]

Originally Posted by KRSW

Here is a video I recommend to you.
https://youtu.be/b2qimyrNK7M

LOL. The channel name says it all there.

[jcard]

Originally Posted by KRSW

Here is a video I recommend to you.
https://youtu.be/b2qimyrNK7M

The different ways in which the advanced stats could be generated is a critical determinant in how they can be used, even for those which are produced a posteriori (e.g., the Statcast data).

For example, if exit velocity is determined by a value distribution based on the outcome of the play, then it is basically flavor text. Conversely, if it is produced by a similar distribution centered instead on the relevant player attribute rating, it can be used as an “truer” insight into that rating than offered by the traditional stats (by, for example offering evidence that the player is hitting into good or bad luck).

BABIP, in particular among more traditional stats, is a similar case for both batters and pitchers.

[MathBandit]

Originally Posted by jcard

The different ways in which the advanced stats could be generated is a critical determinant in how they can be used, even for those which are produced a posteriori (e.g., the Statcast data).

For example, if exit velocity is determined by a value distribution based on the outcome of the play, then it is basically flavor text. Conversely, if it is produced by a similar distribution centered instead on the relevant player attribute rating, it can be used as an “truer” insight into that rating than offered by the traditional stats (by, for example offering evidence that the player is hitting into good or bad luck).

BABIP, in particular among more traditional stats, is a similar case for both batters and pitchers.

This is also my understanding. That some values (like specifically EV) are pure flavour and 'useless' in OOTP. In particular I find that particularly frustrating for hitters- since for pitchers I can basically get everything I need from a combination of FIP, xFIP, and K%-BB%, while for hitters there's essentially nothing particularly predictive other than K% and BB% which are both a much smaller part of the puzzle for hitters (who do have an impact on BIP quality) than for pitchers (who don't); if my 33-year old slugging 1B has his HRs and BABIP both drop heavily in the first 2 months of a season while K% and BB% mostly stay about the same, there is basically no way for me to know if its because he's washed or because he is getting unlikely (and the inverse is true if someone my scout says is an AAAA-type starts mashing for a few months)- whereas for pitchers we have process stats so I can just look at their K%-BB% and know how sustainable it will be.

[Charlie Hough]

Originally Posted by OutS|der

League totals Matter more than ratings. The game is driven by league totals with the help of the modifiers to adjust based on ratings.

No, league totals directly affect the calculated statistical output, but they are not what drives it. Player ratings drive results and outcomes during games, at bat by at bat, and the league totals and modifiers help ensure that those results are calculated and adjusted on a macro level to ensure compliance with the frequency and ratios of events for a given historical year, an era, or whatever conditions you've chosen for your league.

As the OOTP manual has stated for years:

League totals "are simply a basis for calculation, which ensures the ratios of these events remain accurate compared to real life."

[OutS|der]

Originally Posted by Charlie Hough

No, league totals directly affect the calculated statistical output, but they are not what drives it. Player ratings drive results and outcomes during games, at bat by at bat, and the league totals and modifiers help ensure that those results are calculated and adjusted on a macro level to ensure compliance with the frequency and ratios of events for a given historical year, an era, or whatever conditions you've chosen for your league.

As the OOTP manual has stated for years:

League totals "are simply a basis for calculation, which ensures the ratios of these events remain accurate compared to real life."

Meaning 250 power rating is just a number until you add the league totals and modifiers. With one set of totals and modifiers a player can hit 70+, another and he only hits 30.

[Charlie Hough]

Originally Posted by OutS|der

Meaning 250 power rating is just a number until you add the league totals and modifiers. With one set of totals and modifiers a player can hit 70+, another and he only hits 30.

Yes, but without the ratings, literally nothing happens. If you take away the ratings, literally no game results would be generated. You would remove the basis for how they're generated. If you take away the league totals and modifiers, then you would certainly get results, and they would all be based on the modern MLB default, which leaves the pure ratings-based calculations and results generation as the game engine generates them, with no modification of the underlying probability conditions.

Now, we can certainly agree that the ratings, league totals, and modifiers are all working together. But I certainly wouldn't say that the league totals matter more than player ratings. It's that kind of idea that sometimes drives people away from OOTP because they come to believe that player ratings don't really matter, and outcomes are actually being generated and forced by the league totals. That's not what's happening.

I just wanted to make it clear what's really going on, because there has been a lot of mystery, confusion, and misunderstanding about this over the years. I didn't even know how it worked for a long time, until I read the right manual section and got some clarification from the development team.

[jcard]

Originally Posted by krownroyal83

I didn’t mean for this to turn into an argument. I basically wanted to know if there is a clutch factor in the playoffs. The smart, long term players of the game chimed in and have answered my question. There is no clutch factor. That’s all I needed to know. For immersion sake I’m gonna pretend there is. Makes for good story lines.

I will apologize for my role in your question getting overshadowed by disputatious exchanges— ironically, the result of unartful attempt to advocate the legitimacy of the question.

[jcard]

Originally Posted by Charlie Hough

Yes, but without the ratings, literally nothing happens. If you take away the ratings, literally no game results would be generated. You would remove the basis for how they're generated. If you take away the league totals and modifiers, then you would certainly get results, and they would all be based on the modern MLB default, which leaves the pure ratings-based calculations and results generation as the game engine generates them, with no modification of the underlying probability conditions.

Now, we can certainly agree that the ratings, league totals, and modifiers are all working together. But I certainly wouldn't say that the league totals matter more than player ratings. It's that kind of idea that sometimes drives people away from OOTP because they come to believe that player ratings don't really matter, and outcomes are actually being generated and forced by the league totals. That's not what's happening.

I just wanted to make it clear what's really going on, because there has been a lot of mystery, confusion, and misunderstanding about this over the years. I didn't even know how it worked for a long time, until I read the right manual section and got some clarification from the development team.

Personally, the most important question is that once modifiers, etc., are calculated, that every game event is determined independent of the previous events’ resolutions (by which I mean, for example, that a league batting average to date below the expected value does not alter the ratings-based probabilities of a hit in the next at bat). My understanding, on the other hand, is that player ratings are covariant to keep the dstribution of ratings stable.

[pauwoo]

Originally Posted by MathBandit

This is part of the reason that I would disagree heavily with the idea that ratings matter more than something like FIP for a pitcher. Partly because of the context of LTMs, and partly because the ratings are from an inaccurate and biased source.

If I want to know if a 6-sided die is balanced, I'm not going to ask a high school maths student to measure each of the sides and provide me with a mathematical analysis of the balance of the die if I can just roll it 500 times instead and see the results.

You’d rather roll it 500 times yourself rather than chill while some high schoolers do the work for you? I say let the kids do it - let them feel useful, builds self esteem. Character, even.

[Dave Stieb II]

Originally Posted by pauwoo

You’d rather roll it 500 times yourself rather than chill while some high schoolers do the work for you? I say let the kids do it - let them feel useful, builds self esteem. Character, even.

And that is wisdom.

[Charlie Hough]

Originally Posted by jcard

Personally, the most important question is that once modifiers, etc., are calculated, that every game event is determined independent of the previous events’ resolutions (by which I mean, for example, that a league batting average to date below the expected value does not alter the ratings-based probabilities of a hit in the next at bat).

Yes, this is exactly what happens. In OOTP, prior statistical outcomes do not alter the ratings-based probabilities that certain events will occur in future at-bats. Unfortunately, that is what some hardcore baseball sims do. They start dramatically modifying the probabilities for future events in order to force individual players to produce extremely accurate stats that are close to their real-life totals for a given season. For example, in some sims, if a player has already hit 20 home runs by July, and he only hit 25 home runs all season in real life, the future probability that he will hit another home run is massively reduced.

Markus never wanted OOTP to do that, so that's why he set up the league totals system, to help ensure that leaguewide statistical output is accurate and that players with the best capabilities to generate specific types of results will tend to be the league leaders, without penalizing them or trying to prevent them from generating more of those results once their individual totals get to certain levels. I think it works well, because it allows true ratings-based statistical probabilities to play out in a fair and legitimate way. It also provides some really fun and interest what-if results on occasion, where players can individually overachieve or underachieve based on their legitimate ratings and the same fundamental ratings-based probability calculations the sim has been using all season. To me, the OOTP approach is more organic and "real" than artificially trying to force "realism."