Injuries and Starting Players

[Gern44]

Originally Posted by Questdog

I don't think there are players who are injury prone in real life (except for a very few cases, like Dave Dravecky trying to pitch when he had cancer or any idiot who uses a head-first slide). I think players seem prone sometimes since the mind likes to find patterns in events and when a player suffers two or three injuries, he is marked in the mind as prone. But in reality, it is probably just random chance being unkind to some (which we would still see, if all players were given the same rating in OOTP).

My opinion (cause you know you want to hear it) is that this probably isnt 100% accurate. I'm sure there's some validity to it, but I'd bet that, overall, people are simply more fragile than others and it manifests itself in more frequent injuries.

Im certain some people have stronger bones... muscles... tendons... everything... than other people, are are less apt to suffer injury than someone else.

The other factor thats missing is recovery time. That certainly varies person to person.

[Questdog]

Well, I'd say an 80-year-old woman and a 22-year-old man would have different injury rates when sliding headfirst into second.....but we're talking all athletes here.......certainly, late 30's players are more prone to injury than early 20's players and pitching produces more injuries than batting, but withing a group of like players, I don't think there are any players significantly more prone than others.....like I said, I think it's only in hindsight that our brains want to find patterns and put prone labels on people......and even if every player had the same injury rating, you're brain would still find patterns of players looking injury prone......but you wouldn't have the players that NEVER make it through a season without a significant injury.....find me a player in real life who was ALWAYS getting hurt.....

I found this article on injury-prone players to avoid in a fantasy league for 2011: 2011 Fantasy Baseball: 10 Injury-Prone Players You Should Avoid in the Draft | Bleacher Report

Nelson Cruz - 159 games in 2012
Jake Peavy - 32 starts in 2012
Carlos Beltran - 151 games in 2012
Chase Utley - 83 games
Carl Pavano - 13 starts
Chipper Jones -114 games
Josh Hamilton - 148 games
Justin Morneau - 134 games

The other two players mentioned were already hurt and never played in 2012 (Sizemore and Brandon Webb)

Of the 8 players only 2, Utley and Pavano, missed significant time and Pavano's looked more like he just stunk than really getting injured....Chipper only really missed about two weeks, which for a 40 year old ain't nothing to complain about in these hangnail puts you on the DL times.....

[VanillaGorilla]

Originally Posted by Questdog

find me a player in real life who was ALWAYS getting hurt.....

Jim Edmonds came to mind. His most similar player on BBREF is Ellis Burks...yeah, he comes to mind also.

Beltran, who is in your list, is the most similar to Burks, but he played 159 games, or more, three times. He doesn't come to mind as a guy you just knew would have an oblique strain or something that would shelve him every year.

Reggie Sanders is another guy that comes to mind as a guy that was great when healthy, but was always getting hurt.

Ron Gant is the most similar player to Sanders. Yeah, he also rung a bell. What is interesting about Gant is that he played in at least 152 games in each of his four seasons before his motorcycle accident which cost him all of the 1994 season and never played in 140 games in any of his 10 seasons after that. I don't think it is merely hindsight and the human tendency to find patterns where none exist to make the observation that Gant became more injury prone following a traumatic accident.

Eddie Murray was less injury prone because of how he played. It was a conscious decision on his part. Rex Hudler played a different type of game in the field and on the bases and, therefore, was more likely to be injured in a game than Murray.

Pete Rose would chide George Foster about needing a new uniform once a month where Rose would need one by the third inning of each game. Two different guys put together in two different ways.

To argue that everyone is the same and that everyone that plays baseball has the same chance of being injured as anyone else (given the age parameters you offer) is, imo, your brain wanting to find a pattern of sameness across the board where it does not actually exist.

[Questdog]

Not one of the players you mentioned had more than two or three seasons in their career where they had a season ending injury that cost them most of the year. In OOTP, players who go out for the year over and over are a lot more commonplace.

The only way to answer the argument would be to do a study not on the average time lost in a career, but on whether a player who missed a big chunk of the season one year is more likely to miss a big chunk of one of the next few seasons or not, disregarding players who missed chunks of consecutive seasons due to the same injury.

[VanillaGorilla]

Originally Posted by Questdog

Not one of the players you mentioned had more than two or three seasons in their career where they had a season ending injury that cost them most of the year. In OOTP, players who go out for the year over and over are a lot more commonplace.

The only way to answer the argument would be to do a study not on the average time lost in a career, but on whether a player who missed a big chunk of the season one year is more likely to miss a big chunk of one of the next few seasons or not, disregarding players who missed chunks of consecutive seasons due to the same injury.

The season ending injury thing is a completely different discussion.

Just as we can base an expectancy of a player performance based on past performance, we can also place an expectancy on a player missing time due to injury based on past injuries.

No GM could reasonably expect to get 162 games out of a 30 year old Jim Edmonds, but it would not have been unreasonable to think that you had a good chance of getting 162 games out of a 30 year-old Cal Ripken. Is this something that could be ascertained by a team physical? No. Just as you couldn't ascertain that Wade Boggs or Tony Gwynn would probably get you 200 hits based on a physical examination, either.

Heck, Ron Gant was an absolute stud of a physical specimen, but after his accident, it should be no surprise that the same in-game jarring and twisting of parts that a 1992 Gant could endure could place the 1996 Gant on the shelf.

In this regard, I think the OOTP injury model is very good. Once injured, a guy should be prone to further injury at a greater weight. To what degree and extent is debatable. But the concept is a realistic one.

[Questdog]

The season ending injury thing is the only thing I have been discussing all along........it is the manifestation of what is wrong with the OOTP injury model....

You cannot prove or disprove my theory by running out ballplayers who have multiple injuries....even if it was a given that no player at any given moment had more of a chance at an injury than any other, you'd still have players who seemed to get more than their share......

Like I said, the only way to prove or disprove it is to do a study that looks FORWARD from an injury, not backwards......

[Charlie Hough]

Originally Posted by VanillaGorilla

In this regard, I think the OOTP injury model is very good. Once injured, a guy should be prone to further injury at a greater weight. To what degree and extent is debatable. But the concept is a realistic one.

This is the entire issue: the concept is realistic. But its execution for historical leagues is a showstopper for some historical gamers and particularly customers that OOTP might like to attract from competing hard-core baseball sims.

This kind of audience is NOT going to accept the idea that injuries can be purely random and can then start accruing to the point of ruining entire career arcs and statistical output when no such injury tendencies existed for these players in real life.

If we're playing fictionals or projecting into an MLB future from 2013, that might make sense. But it will not make sense to users who expect some variation but do not want to see players who rarely suffered real life injuries being injured dozens of times in their OOTP career. Or players having their careers destroyed when the user is otherwise trying to use settings to keep performances close to real life.

Based on what's been discussed here, this appears to be an aspect of OOTP historical gameplay that is in total conflict with its marketed features and performance.

The game is promoted as a highly realistic simulation of baseball with the ability to have historical players perform much like they did in real life. Yes, there are options to allow player development and allow your baseball environment to deviate from real life, but there are also supposed to be options to keep the game results and player careers much closer to real life.

This injury model works against those latter options and conflicts with the game's marketing messages. If I'm a Diamond Mind or SOM guy and I want to use OOTP as a single season replay engine or a 3-year recalc engine for a historical draft league, I do not expect injuries other than to the extent required to prevent players from over-usage or to reflect game engine randomness.

And I certainly do not expect that a player who is injured in the first month of the season might be injured several more times in that same season, thus reducing his at bats by 200 or more and ruining my replay. Multiply that by the dozens of players that are affected by injuries every season and you have quite a mess.

I'm fine with the injury system for a fictional environment IF it reflects real life tendencies per professional baseball player. But that has yet to be established. No one has presented any statistically significant data to prove that the game's injury system is actually realistic when compared to real life on an individual and not a league-wide basis.

That being said, if it's realistic in those terms, then that's fine for fictional games. But I have a major problem with the injury system when applied to historical games because it's actively working against the output that the game is supposed to generate and the features that are marketed as a value for the product.

[VanillaGorilla]

Originally Posted by Questdog

The season ending injury thing is the only thing I have been discussing all along........it is the manifestation of what is wrong with the OOTP injury model....

You cannot prove or disprove my theory by running out ballplayers who have multiple injuries....even if it was a given that no player at any given moment had more of a chance at an injury than any other, you'd still have players who seemed to get more than their share......

Like I said, the only way to prove or disprove it is to do a study that looks FORWARD from an injury, not backwards......

I didn't look up the injury history of the players I mentioned. However, I did go through my current injury list with 200+ players on it. Of those players, 49 are recovering from an injury that is greater than six months (or recovering from an injury/setback that is combined greater than six months).

Of those 49 players, 14 have had a previous injury or injury/setback combo of greater than 6 months.

No player on my current injury list has suffered 3 instances of injury or injury/setback combo greater than six months.

And OOTP is out of whack in regards to the frequency of season ending injuries how?

These results are from a "Very HIgh" injury setting in a 30 team league.

[VanillaGorilla]

Originally Posted by Charlie Hough

This is the entire issue: the concept is realistic. But its execution for historical leagues is a showstopper for some historical gamers and particularly customers that OOTP might like to attract from competing hard-core baseball sims.

This kind of audience is NOT going to accept the idea that injuries can be purely random and can then start accruing to the point of ruining entire career arcs and statistical output when no such injury tendencies existed for these players in real life.

If we're playing fictionals or projecting into an MLB future from 2013, that might make sense. But it will not make sense to users who expect some variation but do not want to see players who rarely suffered real life injuries being injured dozens of times in their OOTP career. Or players having their careers destroyed when the user is otherwise trying to use settings to keep performances close to real life.

Based on what's been discussed here, this appears to be an aspect of OOTP historical gameplay that is in total conflict with its marketed features and performance.

The game is promoted as a highly realistic simulation of baseball with the ability to have historical players perform much like they did in real life. Yes, there are options to allow player development and allow your baseball environment to deviate from real life, but there are also supposed to be options to keep the game results and player careers much closer to real life.

This injury model works against those latter options and conflicts with the game's marketing messages. If I'm a Diamond Mind or SOM guy and I want to use OOTP as a single season replay engine or a 3-year recalc engine for a historical draft league, I do not expect injuries other than to the extent required to prevent players from over-usage or to reflect game engine randomness.

And I certainly do not expect that a player who is injured in the first month of the season might be injured several more times in that same season, thus reducing his at bats by 200 or more and ruining my replay. Multiply that by the dozens of players that are affected by injuries every season and you have quite a mess.

I'm fine with the injury system for a fictional environment IF it reflects real life tendencies per professional baseball player. But that has yet to be established. No one has presented any statistically significant data to prove that the game's injury system is actually realistic when compared to real life on an individual and not a league-wide basis.

That being said, if it's realistic in those terms, then that's fine for fictional games. But I have a major problem with the injury system when applied to historical games because it's actively working against the output that the game is supposed to generate and the features that are marketed as a value for the product.

I totally get what you are saying, but I ask what is the preferred solution?

If Lou Gehrig can have a piano drop on his head in April of 1927 and be out for the season, then someone else takes his playing time that didn't have 1927 playing time. Is that player, who maybe played in 10 games, prone to rate of injury 15 times greater than Gehrig because he played in 1/15th of the games Gehrig did? What you wind up with is players dropping dropping like flies due to their lack of playing time once a guy that played an entire season is felled by injury, if you are basing injury chances on GP.

Once you deviate from reality, then the model will make "unrealistic" deviations from reality, whatever that model is.

Is it unrealistic that Gehrig gets injured at all in 1927? Some might say that it is. I am in the camp that believes that it is unrealistic for Gehrig, or anyone else in 1927 to have known that it was impossible for him to get injured.

I think the only way around this is to use historical line-ups. Anything other than that is not reality and if you know what the deviations from reality are going to be ahead of time, then these deviations are scripted as opposed to random.

[VanillaGorilla]

Originally Posted by Questdog

You cannot prove or disprove my theory by running out ballplayers who have multiple injuries....even if it was a given that no player at any given moment had more of a chance at an injury than any other, you'd still have players who seemed to get more than their share......

You cannot disprove my theory that no player has a greater chance of hitting a home run than any other player in a given moment by running out names of players who hit lots of home runs. Even if it was given that no player had more of a chance to hit a home run than any other, you'd still have players who seem to get more than their share.

[Charlie Hough]

Originally Posted by VanillaGorilla

I think the only way around this is to use historical line-ups. Anything other than that is not reality and if you know what the deviations from reality are going to be ahead of time, then these deviations are scripted as opposed to random.

What we have here is a fallacy of a false dilemma. It's not all-or-nothing. It's not all scripted or all random.

All it would take is some programming to allow the user to:

1. See the hidden injury information
2. Adjust the injury model to increase or reduce the chances of subsequent injuries, set limits on injuries, use historical stats as a basis for injury proneness and chances, etc.

This is no different than offering options to regulate and tweak all the other features in the game. And there are many other baseball sims that have used injury models based on historical player statistics. This has been a standard practice in the industry since long before Markus was born.

So I'm hardly asking for anything radical. Personally, I'm comfortable with some randomness when it comes to historical player injuries, but I reject any notion that it's acceptable for OOTP's injury model to blatantly ignore historical reality and impose its randomness and its own mathematical model to an extreme that significantly undermines historical integrity and accuracy.

Obviously the tolerance level for randomness is subjective in the end, but that doesn't mean that there is no gray area and that it's an either/or proposition.

[ihatenames]

Originally Posted by Questdog

I don't think there are any players significantly more prone than others.....like I said, I think it's only in hindsight that our brains want to find patterns and put prone labels on people......and even if every player had the same injury rating, you're brain would still find patterns of players looking injury prone......but you wouldn't have the players that NEVER make it through a season without a significant injury.....find me a player in real life who was ALWAYS getting hurt.....

A few off the top of my head:

Josh Beckett- Only has pitched over 200 innings 3 times. Only started 30 or more games 4 times in his 13 year career.

Kerry Wood- what more needs to be said

Jake Peavy - had a period during 2008-2011 where he had significant problems staying on the field. Also struggled with a few injuries earlier in his career

Scott Rolen- from 2002-2012 only played a 150 games or more once. And that was 2003. The next highest totals were 142, 133 and 115.

Rickie Weeks- Only played over 130 games once in his career.

Jose Tabata- Things aren't looking good for him in the early stages of his career as he has yet to complete a season without missing significant time.


This doesn't address your claim that season ending injuries in OOTP or modeled incorrectly, but clearly there are some players who get injured more than others. Physcially we are not all created equal Either through genetics, bone density or strength of tendons and countless other subtle differences.

As a fictional player I think OOTP does a pretty good job in this department and it shows on the career leaderboards.

My 2 cents

[VanillaGorilla]

Originally Posted by Charlie Hough

What we have here is a fallacy of a false dilemma. It's not all-or-nothing. It's not all scripted or all random.

I offered no false choice all or nothing proposal. I asked how you would model things differently.

Originally Posted by Charlie Hough

All it would take is some programming to allow the user to:

1. See the hidden injury information
2. Adjust the injury model to increase or reduce the chances of subsequent injuries, set limits on injuries, use historical stats as a basis for injury proneness and chances, etc.

And here you did, and I do like this.

To your previous comparison of OOTP to SOM, I think OOTP's model is superior because in SOM Gehrig or Ripken, or any player that played 162 games will not have a piano fall in their head (or take a fastball to the eye socket). I understand that some don't like this, and that is fine.

Originally Posted by Charlie Hough

This is no different than offering options to regulate and tweak all the other features in the game. And there are many other baseball sims that have used injury models based on historical player statistics. This has been a standard practice in the industry since long before Markus was born.

Agreed that mo options is mo better. As I said above, I think Marcus has a model that is superior to SOM. Was Tony Conigliaro injury prone, or was his his career shortened by a freak event? If Wally Pipp doesn't take that day off, does Lou Gehrig's most similar player become Gordy Coleman? OOTP gives us the playout of these possibilities. SOM does not.

Originally Posted by Charlie Hough

So I'm hardly asking for anything radical. Personally, I'm comfortable with some randomness when it comes to historical player injuries, but I reject any notion that it's acceptable for OOTP's injury model to blatantly ignore historical reality and impose its randomness and its own mathematical model to an extreme that significantly undermines historical integrity and accuracy.

Obviously the tolerance level for randomness is subjective in the end, but that doesn't mean that there is no gray area and that it's an either/or proposition.

This is key. One random piano dropping on Gehrig's head results in a history instantly radically different from history. Even if the odds in the code were one in a million, when it did happen, the person experiencing it might cry foul and say it should NEVER happen because it didn't happen in real life.

If we want history to be altered so we can take the Cubs to the World Series or get 30 wins from Sandy Koufax, something has to change.

If we want Tony Conigliaro to play 20 years in a realistic manner, we have to allow for Gehrig and Ripken to have careers that don't get them into the HOF.

[SunDevil]

Originally Posted by charlie hough

what we have here is a fallacy of a false dilemma. It's not all-or-nothing. It's not all scripted or all random.

All it would take is some programming to allow the user to:

1. See the hidden injury information
2. Adjust the injury model to increase or reduce the chances of subsequent injuries, set limits on injuries, use historical stats as a basis for injury proneness and chances, etc.

This is no different than offering options to regulate and tweak all the other features in the game. And there are many other baseball sims that have used injury models based on historical player statistics. This has been a standard practice in the industry since long before markus was born.

So i'm hardly asking for anything radical. Personally, i'm comfortable with some randomness when it comes to historical player injuries, but i reject any notion that it's acceptable for ootp's injury model to blatantly ignore historical reality and impose its randomness and its own mathematical model to an extreme that significantly undermines historical integrity and accuracy.

Obviously the tolerance level for randomness is subjective in the end, but that doesn't mean that there is no gray area and that it's an either/or proposition.

+1

[Charlie Hough]

Originally Posted by VanillaGorilla

This is key. One random piano dropping on Gehrig's head results in a history instantly radically different from history. Even if the odds in the code were one in a million, when it did happen, the person experiencing it might cry foul and say it should NEVER happen because it didn't happen in real life.

Yes, which is why the idea is to have options so the users who want it NEVER to happen can have what they want, and those who are comfortable with the occasional, random piano dropping on Gehrig's head can have what they want.

The problem, as I see it, is that OOTP is dropping pianos on many players' heads and doing it repeatedly over their careers. This is an exaggeration, but hopefully people get the point.

As I'm now examining my historical leagues more closely for this phenomenon, I'm seeing many players who have clearly gone into catastrophic, recurring injury mode in just a few years of the league. All you need to do is look at their player histories, and you can see the repeated attacks of injuries that are a product of the OOTP snowball effect.

Now when I examine trades, I look at the player histories more closely, and I'm seeing many players who are doomed. I know that they're going to keep getting hurt, even if OOTP says they are durable. And as they keep getting hurt, presumably the frequency and severity will keep increasing. It's rendering a lot of players devoid of value or making them far too risky to acquire. By this I mean that it's more extreme than what you'd expect based on real life occurrences. Sure, there will always be guys with injury histories that you want to avoid, but should there be this many?

I am convinced that OOTP is taking this much too far and that these recurrence rates and increased chances are far higher than they are in real life. The problem is that I don't think there are any data sets that we could use for real life comparison. Certainly the chance data can't be out there. Whatever those percentages are in OOTP, they're pure conjecture and fantasy.

But I'll be damned if I can find as many players in real life that are suffering 6, 8, and 10 injuries in just a few seasons as what I'm seeing in OOTP. Sure, it happens on occasion, but I don't recall it happening to as many MLB players in a span of three years.

And, remember folks, the exceptions prove the rule. If you have to do a lot of work to find the 10 or 20 guys that have been plagued with injuries like this over the past 10 years, then you've already proven the point: the numbers of players in OOTP that are going through this seems significantly larger and is happening over a much shorter range of time. So it seems awfully unrealistic.

[Gomer]

With all this injury talk.... do injuries affect talent/ratings or not?

I got an answer in one thread that said "yes".... I've gotten answers from others that said "no".

What is the real answer?

[Questdog]

The real answer is sometimes......players can take hits to their ratings due to injury, but don't always suffer.....

[David Watts]

Maybe just like there is a Talent Change Randomness number that we can raise or lower to suit our fancies, there should be an injury randomness number. If set low, players could get hurt, but little or no effect would be seen to the players durability or playing ability. He would just miss the games.

[Charlie Hough]

Originally Posted by David Watts

Maybe just like there is a Talent Change Randomness number that we can raise or lower to suit our fancies, there should be an injury randomness number. If set low, players could get hurt, but little or no effect would be seen to the players durability or playing ability. He would just miss the games.

Excellent idea and heading in the right direction.

[SteveP]

Originally Posted by Charlie Hough

As I'm now examining my historical leagues more closely for this phenomenon, I'm seeing many players who have clearly gone into catastrophic, recurring injury mode in just a few years of the league. All you need to do is look at their player histories, and you can see the repeated attacks of injuries that are a product of the OOTP snowball effect.

Now when I examine trades, I look at the player histories more closely, and I'm seeing many players who are doomed. I know that they're going to keep getting hurt, even if OOTP says they are durable. And as they keep getting hurt, presumably the frequency and severity will keep increasing. It's rendering a lot of players devoid of value or making them far too risky to acquire. By this I mean that it's more extreme than what you'd expect based on real life occurrences. Sure, there will always be guys with injury histories that you want to avoid, but should there be this many?

I am convinced that OOTP is taking this much too far and that these recurrence rates and increased chances are far higher than they are in real life. The problem is that I don't think there are any data sets that we could use for real life comparison. Certainly the chance data can't be out there. Whatever those percentages are in OOTP, they're pure conjecture and fantasy.

But I'll be damned if I can find as many players in real life that are suffering 6, 8, and 10 injuries in just a few seasons as what I'm seeing in OOTP. Sure, it happens on occasion, but I don't recall it happening to as many MLB players in a span of three years.

And, remember folks, the exceptions prove the rule. If you have to do a lot of work to find the 10 or 20 guys that have been plagued with injuries like this over the past 10 years, then you've already proven the point: the numbers of players in OOTP that are going through this seems significantly larger and is happening over a much shorter range of time. So it seems awfully unrealistic.

I highly recommend that you start a thread in the Beta Forum with the focus on what you've said here (as I think it is the best explanation so far in this thread). In that Forum, you will find people who are very knowledgeable about how OOTP handles injuries, and you have a better change of attracting Markus' attention. I can see a situation here where Markus might readily tweak some existing formula, if it's just a matter of dialing down something.

No disrespect intended for the folks on this forum. I just think this specific issue is a good candidate to raise in that other forum as well.