Injuries and talent/ratings loss

[mbi]

Originally Posted by Getch

I think then you'll find that there are way too many torn ACL's and massive ligament damage than there are in real-life. You're just hiding the inherent problem.

Can't you just adjust injury frequency? Then again I really feel injuries are fine in the game, so I guess I'm biased.

[Syd Thrift]

To those who think that torn ACLs and so on happen too often in the game, by all means run some simmed seasons and compare to actual data. Two reasons for this:

1. Things that "seem to be correct" often are not, and
2. The best way to convince Markus that there is a problem is by presenting him with data. Anecdotes are not data.

[Getch]

As Steve said, the issue is not the # of injuries, nor the types of injuries. The issue (for me) is the effects of the injuries, specifically for established major leaguers in their prime.

And as Steve said, I know anecdotal or "in my league" is not solid evidence. Unfortunately I do not have the resources to run a test to check my theory. I guess the test would be how often players who've established themselves (3+ seasons of good production) then lose their ability before the age of, say, 32.

However, another test would just be to look at the OOTP code. Does it take account the player's age/position/injury? If it does not, well, then, we have a problem. Unless people are challenging my assumption that injury/age have an effect on how much the player rebounds from an injury. If it does take those factors into account, then we need to run an analysis against real life, as mentioned above.

[CBL-Commish]

Originally Posted by Getch

And as Steve said, I know anecdotal or "in my league" is not solid evidence. Unfortunately I do not have the resources to run a test to check my theory. I guess the test would be how often players who've established themselves (3+ seasons of good production) then lose their ability before the age of, say, 32.

I think you need to compare that to real life. Look on bb-ref at players who are currently ~32, born in 1977. Sure, there are guys like Carlos Beltran, Brian Roberts and Roy Halladay who are still more-or-less in their primes. But the list also includes Eric Chavez, Andruw Jones, Ty Wigginton, Lyle Overbay, Jay Gibbons, D'Angelo Jimenez, Brad Wilkerson, Danys Baez, Bruce Chen, Adam Eaton, Kerry Wood... There has to be two or three times as many guys who're in steep decline or out of the league than still in their primes at 32. Many or most of them accelerated out the door by injury.

[Getch]

Our league brought injury log into this discussion a few months back, when we had a lot of owners upset over the issues I laid out. This is what he said:

There are some misconceptions in that thread about how often real life injuries occur - whoever thinks no MLB team loses their front three starters should look at what happened to the Blue Jays this year - I think they had their top six SPs on the DL early in the season. And Oakland lost four or five players (don't remember which) to injury in a single game. That's not to say OOTP has injury numbers right -- injuries are far too long to batters in OOTP, for example -- but real life baseball is murder on players, and managing around injuries is among the most important tasks a real life GM has. Whether OOTP GMs want that challenge is another matter.

The biggest complaint in that thread, and it's my biggest complaint about OOTP's injury model as well, is how player ratings are affected by injury. Back in OOTP2007, players lost potential instantaneously; one day they were a top prospect, and the next they were junk. Users complained about this, saying that it made no sense to see these unexplained sudden drops in potential. On the forums, many users said that injuries were the only real life reason a player might go from being a top prospect to a dud overnight. That might sound logical, but unfortunately they had no evidence to support any of this, and they screamed loud enough that Markus went ahead and removed most random talent drops and replaced them with injury-triggered talent drops. And the thing is, for OOTP's development mechanism to work realistically, there need to be a lot of talent drops, which meant that a ton of injuries needed to produce ratings hits; even day-to-day injuries caused talent loss in OOTP9. The irony of all of this is that Markus actually fixed the problem of instantaneous potential changes - he made them occur gradually - so the original reason injuries became the trigger for potential drops doesn't even exist any more.

In any case, I did a study on how real life player performance is affected by injury and sent it to Markus last year; as a result, he toned down the effects of injury, and particularly of shorter injuries, but didn't come particularly close to modeling real life - I believe he thinks users want injuries, rather than randomness, to cause talent loss. In any case, I looked at every player who was out for 60 days or more in the 2007 MLB season, compiled their stats from the 18 months before and after injury, and translated into OOTP ratings. From that sample, this is what seems to be true:

-real life baseball players decline on average by about 2-8%, in OOTP ratings terms - the drop depends on the rating. That is, even long injuries aren't normally debilitating, and shorter injuries really should have no effect on player performance;

-the 'catastrophic' injury -- i.e. one that turns a big leaguer into a minor leaguer -- pretty much doesn't happen to batters in real life, with only rare exceptions. It does seem to happen to about 10% of pitchers, though;

-for batters, the one stat that declines dramatically is stolen base attempt rate following a serious leg injury -- SB rate falls by about 50% on average, and dropped dramatically in every instance, no exceptions. SB success doesn't change at all though;

-for pitchers, as is often discussed in sports injury circles, elbow injuries did have a serious effect on control (10% or so in OOTP terms), and shoulder injuries did have a serious effect on K rate (again, around 10% on the OOTP Stuff scale). Finger injuries seemed to affect HR rate (Movement, in OOTP terms), though there were so few data points there it's really anyone's guess if that's just a fluke or something more meaningful.

I was planning on finishing the study off (at the time I did it, I had incomplete data on some pitchers, since at the time a few had only recently returned from injury) and then posting it to beta. If there were enough voices on beta pushing for a change, I imagine Markus could be persuaded to make this more in line with real life.

All the best,
Ian

Ian did break injuries down by batter vs pitcher, although I do not think he did by age. However, he did a study and the numbers are in-line with some of my complaints.

[thomamon]

Originally Posted by battists

I find myself wondering if the phrase "works as intended" has been taken slightly out of context. That's not an accusation, but I suspect that what Markus may have been talking about is the injury system in general. The context above makes it sound like Markus has deliberately coded the game to injure star players, which I absolutely do not think is the case.

I don't think that anyone actually thinks that is coded to hurt star players, I don't think it was meant that way at all.

Originally Posted by battists

If anything, I'm more concerned by the "talent hits ruin the player afterwards" aspect of this conversation. We're about of our time in the beta cycle, but I'll make sure this gets a last look, at any rate...

I think everyone in this thread can agree that this is what everyone will want! If this gets a look at, everyone will be happy.

[mbi]

Honestly I don't know much about the injury model (I accept and move on), but some type of system where a player has some injury, train reports, then an outside source (another trainer/league trainer/etc) says "Hey this rib strain appears to not be a big deal, but x% of people who get them have long term effects if they do x" Then you make a decision and results occur close to what those percentages are.

[Qrusher14242]

i just had a guy at 28, have these injuries:

strained back 2-3 weeks
sprained ankle 2 weeks
hand contusion 2 days

his ratings before the injuries:

81 contact 100 power 86 eye

after injuries

44 contact 96 power 71 eye

Is that normal? he lost like almost 80 points in avg and like 60 points in VORP.

[battists]

One other thing to point out, that may go back to Markus' original intent:

Players MUST lose talent in OOTP. If they didn't, the leagues would grow more and more talented over time. Every player would just keep getting better and better.

So, in the engine, there HAVE to be checks and balances. Much of that IS associated with injuries and aging, which is, I think, largely how it should be. (This is compared to previous versions like OOTP 6 wherein it was almost entirely random.) I think there is room for continued refinement in that algorithm. (There's a conversation going on over in the beta forums about this right now.)

I DO know for a fact that at LEAST the impact of day-to-day injuries on ratings has either been eliminated entirely or drastically reduced. And perhaps we will find that with OOTP 11, that was enough to move things from "unacceptable" in the minds of people in this thread to "acceptable."

But, I can't claim to have run any tests to prove those results one way or the other. Unfortunately, as Getch intimated, it's not an easy thing to test at a statistically-significant level.

[Getch]

Originally Posted by CBL-Commish

I think you need to compare that to real life. Look on bb-ref at players who are currently ~32, born in 1977. Sure, there are guys like Carlos Beltran, Brian Roberts and Roy Halladay who are still more-or-less in their primes. But the list also includes Eric Chavez, Andruw Jones, Ty Wigginton, Lyle Overbay, Jay Gibbons, D'Angelo Jimenez, Brad Wilkerson, Danys Baez, Bruce Chen, Adam Eaton, Kerry Wood... There has to be two or three times as many guys who're in steep decline or out of the league than still in their primes at 32. Many or most of them accelerated out the door by injury.

There's a big difference on your list. Of the batters:

Chavez: Tough to say since he's been constantly injured we do not know how good he is now.

Jones: I'll give you him

Wiggington: Only 2 seasons (2006/2008) of an OPS over .800 (not even sure you can say he declined yet, as one season is not enough sample size)

Overbay: Only had one good season (2006: .312/22 HR, and one ok season .301/16 HR)

Gibbons: Only one season of over .800 OPS (2005)

Jimenez: When was he good?

Wilkerson: Never had a BA over .270, only had one season with over 20 HR's.

With the exception of Jones, none of these guys were All-Stars. They may have had one or two good seasons, but that's it. We see this fluctuation in OOTP too, where an everyday player has a random good season now and then (I love that about OOTP).

Anyway, the point is, my beef is with All-Stars losing it. These are average players getting slightly worse, and since there are a lot of average players, a below-average player is going to be cut quickly. The only case I can come up with in the last 5-10 years in Jones, yet in league after league it's a common occurrence to see batters go from All-stars to backups before they hit 30.

As for the pitchers, I am willing to be more understanding of injuries. But you again did not list any sure-fire All-star, but more everyday pitchers, or pitchers with high potential but did not reach that potential

[Syd Thrift]

I *do* wish the game would skew injuries away from position players and towards pitchers. I put together data on that a year and a half ago and got nothin'. I also saw that while certain positions (2nd, catcher) IRL get hurt more, this doesn't happen in the game so much. Injuries were closer in least year's version but not exact; I still ended up using the default injury file instead of making my own (IIRC, I eliminated long-term baserunning injuries altogether, which isn't realistic on a game to game basis but forced the engine to disregard a lot of the injuries that happened when I ramped things up to High; since pitchers rarely get involved in baserunning plays, they continued to get hurt as often as they would otherwise... this sort of worked).

What I'd really like to see is some control over how often injuries are allotted through that injury.txt file. Pretty much the only way you can turn off CEIs, for example, is by turning *all* of them off for a given injury type. If you have even a single boolean set to allow CEIs for (for example) baserunning, the game will generate exactly as many as before but they will all be of that particular injury.

[Getch]

Originally Posted by battists

One other thing to point out, that may go back to Markus' original intent:

Players MUST lose talent in OOTP. If they didn't, the leagues would grow more and more talented over time. Every player would just keep getting better and better.

But talent at the ML level? I understand prospects losing talent and ratings, and I would think so does everyone else. They don't pan out. If the game wants to use injuries to simulate prospects not panning out, then that's fine by me.

But does OOTP also need *ML* talent to dwindle for players in their primes as well? I would have expected most talent to have been weeded out in the minor leagues, and possibly with unproven rookies.

[TGH-Adfabre]

Considering the customer service experiences I have had in my life, it is impressive how OOTP Developments handles their customers.

[battists]

Originally Posted by Getch

But talent at the ML level?

I can't really think of a compelling reason why NOT at the ML level. I mean, your points are well-taken, and I would value Injury Log's studies over my poor conjecture any day. Oddly given my role, I don't follow baseball all that closely, but there HAVE to be examples of ML (even All-Star caliber) talent who degrades over time, sometimes even rapidly. (Ortiz?)

How much of that is attributable to age? To lack of steroids? To the cumulative effect of injuries? I certainly can't say.

Getch, would you agree that your list of injured players above (your OOTP examples) would be a lot less jarring if it only consisted of cases where the player was SERIOUSLY injured? I mean, a fractured eye socket, I can see a player struggling to come back from that, whether he was an All-Star or not. A 3-day injury or shoulder inflammation, though, do seem "unreasonable." (And potentially have already been addressed, per my earlier note.)

[Getch]

Originally Posted by battists

Getch, would you agree that your list of injured players above (your OOTP examples) would be a lot less jarring if it only consisted of cases where the player was SERIOUSLY injured? I mean, a fractured eye socket, I can see a player struggling to come back from that, whether he was an All-Star or not. A 3-day injury or shoulder inflammation, though, do seem "unreasonable." (And potentially have already been addressed, per my earlier note.)

If I knew OOTP used injury severity as a factor in determining talent/rating loss, then yes. However, you're just moving the issue from one thing to another. Then the issue becomes the frequency of such injuries. If OOTP were to maintain the status quo, then it would have to jack up these severe injuries higher than real-life rates (assuming my hypothesis is correct).

Also, I don't discount the fact that injuries might take its toll on an All-Star. I can see an All-Star getting injured, and just becoming a really good player. Or a really good player becoming a decent player. It's the All-star to backup, or really good player to minor leaguer, that should be extremely rare for players 30 or younger.

[kq76]

The seemingly obvious answer to me is to just have less great players make the majors.

Just comparing a recent league I joined to leagues I'm normally used to, this new league has an unreal amount of talent. The average major league offensive rating seems to be a 7 whereas I'm usually used to it being 6, but more importantly it seems like almost every team has almost a full roster of stars. If these managers who are seeing these numerous injuries to all-stars are in leagues like this one then I'm not surprised that they're seeing this. As TribeFan pointed out above, talent should be close to zero sum so if you have a lot of great players coming up then a lot of established greats are going to have to take a hit too. Regardless, maybe Markus could do something more along these lines, but shouldn't the dev and aging speed modifiers do something for this as well?

Personally, I'm content with how things are in both my solo and regular online leagues so I'm just thinking there's probably something particular about some leagues that is causing this to be a bigger problem than what we might otherwise see from test leagues.

[BMD]

Let's not forget that in real life some pitchers missing upward of a year for Tommy John surgery (major injury) are coming back throwing harder than they did before the procedure. Second thought, forget I brought it up. The water might be muddy enough already.

[Markus Heinsohn]

First, let's not talk about injury frequency again, that one has been brought up enough and it is proven that OOTP's injury frequency is realistic, both in length and how often these lengths appear.

Second, players in OOTP lose talent because of either a) injury, b) randomness or c) aging. Now, balancing these three is the problematic part. I personally feel that OOTP 10 did a good job, it felt realistic for me, and long-term tests proved that overall talent distribution and average talent level remained very stable over 100+ year sims, just as it should be.

The problem is that this part of the game will never work perfectly fine for everyone's taste. That's simply because of perception, and not because of scientifically proven facts, those simply do not exist...

Anyway, in OOTP 11 this part has been tweaked slightly so that talent loss because of injury is a bit less likely.

[Markus Heinsohn]

Originally Posted by Getch

If I knew OOTP used injury severity as a factor in determining talent/rating loss, then yes.

It does, of course.

[Markus Heinsohn]

Originally Posted by Qrusher14242

i just had a guy at 28, have these injuries:

strained back 2-3 weeks
sprained ankle 2 weeks
hand contusion 2 days

his ratings before the injuries:

81 contact 100 power 86 eye

after injuries

44 contact 96 power 71 eye

Is that normal? he lost like almost 80 points in avg and like 60 points in VORP.

That case was possible in OOTP 10, but EXTREMELY rare. The strained back caused this.

In OOTP 11 injuries less than 4 weeks of length do not cause talent drops.