TEAM TRAINERS, WHAT ARE THEY GOOD FOR?All right. I need to be straight with you here: what follows is only part of the picture. After finishing the work, I’ve had some conversation with the developers and I think there are some things about this that I can’t share. That said, what I’m working with here are things that are available to the general public simply by compiling data from the outside (admittedly only with some sweat-equity involved, but still available), and what I’m seeing is helpful in that it strongly suggests that, yes: Your OOTP team trainer does make some small, but identifiable difference in the health of your baseball team.
[Aside: The adroit of you will see that italicized part as the TLDR take-away.]
Off Topic
The Method:
What follows is an OOTP 21 study, but my guess is that the basic nature of trainers has not changed over the versions.
I set up a league of 40 teams in divisions of 10. I then hand-edited all the trainers such that:
Division 1: Good Trainers (all parameters set at 200) focused on Prevention
Division 2: Good Trainers focused on Recovery
Division 3: Quacks (all parameters set at 1) focused on Prevention
Division 4: Quacks focused on Recovery
I then edited all the other available doctors to be Quacks (so the AI wouldn’t replace the current quacks with better players.
I also turned off some things like suspensions and trading just for general purposes.
Injury settings were default, out of the box classic…which is high. I think we use the next level down. Regardless, the point of the exercise still stands.
I then ran a year, took the league injury report, and used a simple perl script to parse the data in the way I parse our BBA data. At this point I could start to categorize a few things.
What follows is an OOTP 21 study, but my guess is that the basic nature of trainers has not changed over the versions.
I set up a league of 40 teams in divisions of 10. I then hand-edited all the trainers such that:
Division 1: Good Trainers (all parameters set at 200) focused on Prevention
Division 2: Good Trainers focused on Recovery
Division 3: Quacks (all parameters set at 1) focused on Prevention
Division 4: Quacks focused on Recovery
I then edited all the other available doctors to be Quacks (so the AI wouldn’t replace the current quacks with better players.
I also turned off some things like suspensions and trading just for general purposes.
Injury settings were default, out of the box classic…which is high. I think we use the next level down. Regardless, the point of the exercise still stands.
I then ran a year, took the league injury report, and used a simple perl script to parse the data in the way I parse our BBA data. At this point I could start to categorize a few things.
I should start by saying that the data is one season’s worth. It also is taken with no attempt to normalize players (so there’s always the chance that distribution of “wrecked” or “fragile” players could be influencing the numbers. That said, each division is represented by 10 teams of 25 players, or 250 total players at any one time: 1,000 total players combined.
Anyway…
For the year, there were a total of 1,876 injuries in this 40-team league. That’s 47 per team, split as follows:
- DTD: 932
- Lost Time: 893
- Setbacks: 51
When I split the data by teams with “Good” Trainers and teams with “Quacks”, the information is pretty clear that the better doctors serve to reduce the total quantity of lost-time injuries (about 20% fewer), as well as reduce the time a player sits on the shelf (about 5% quicker recover).
- TRAINER-TYPE-SUMMARY.PNG (9.54 KiB) Viewed 1310 times
I note that DTD injuries did not seem to be any less prevalent, but recovery may or may not be better under a better doctor. This makes as much logical sense as any, I suppose. But really, I need to run more data to say anything there. In addition, the set-back recovery data is interesting, but a lot of it was based on one or two pitchers who had big set-backs, so I think there’s data quality issues in there.
A few “remember” caveats: First, remember that the injury setting in this run was higher than we use. Second, remember that the gap between “Quacks” and “Good” was the full scale of 1-200. Both these factors will make the impact of trainers we see considerably watered down.
For example, if this league saw five more injuries per team due to their bad doctor, if one were to split that 1-200 range into five separate buckets, you’d say a 40-point gap would be worth one injury. If we were to guess that most of our doctors are between 80-160 in prevention settings, that would say we’d expect a BBA doctor to be worth 1-2 injuries, reduced again by maybe half due to our lower setting, so, maybe our doctors can save us an injury or so every year or two.
Valuable, but not a major deal …unless, of course, that one injury is to your star pitcher.
Similarly, it appears a great doctor can get your guys back on the field a just a touch earlier, but that “greatness” is spread over 1-200 points, so we’re talking about some pretty small numbers once we allocate them.
BOTTOM LINE:
1) The base quality of your trainer does make a difference
2) The difference is measurable, but not massive
Personally, I like that. Without any true data to say why, it feels like a good balance for game play. I should also note that, at the end of the day, this makes me happy too because--while we've all probably told ourselves at one point or another that we assume something matters just because if it didn't it wouldn't be in the game--I personally like to know that whatever effort I'm putting into these decisions actually does make some difference beyond the joy it gives me inside my own mind. Knowing trainers do seem to matter, even if that difference is very small (or particularly because it's very small) makes me happy because it suggest that all the other things I spend sweat on also matter.
Just a little, anyway.
Which is all I want.
Your mileage may vary, but in this instance, OOTP has, once again, made me happy.
The question for us as GMS, of course, is--assuming we agree there's value in a trainer--how much are we willing to pay for that value? Such are the things they pay us the big fake bucks for.
PREVENTATIVE Vs. RECOVERY FOCUS
This is the area I need to step back from. It turns out that the data I have for this cut is kind of wonky and if I present it, will result in some misleading evaluations. Based on feedback from Matt in particular, I don’t think the study I’ve done will assess this well—so I need to think more about it. There’s also the idea that I may need more data.
So, anyway, I’m not going to touch it here.
My general comment at this point is that unless I get time to study it further, I’m going to ignore that aspect of the trainer’s ratings in my own operations. I can’t say that’s the “right” thing to do, but it seems prudent at present.
