An experiment in training a catcher

[allenciox]

So five seasons ago, I bought 2019 AS Carlos Santana (89) and have been working on training him as a catcher --- the vanilla guy has catcher ability of 28 (catcher arm=48), but no position rating at Catcher.

During that time, I have been platooning him against righties with Carlton Fisk (89, Catcher DEF=67, c ability=64, c arm=62) as sub and also going against lefties. This is a great opportunity to learn:

1. What the process is to train a player to be Catcher
2. How important catcher defense is to outcomes in the game when a really bad catcher is used; as catcher defense somewhat surprisingly (at last according to WAR) does not appear to be all that important.

This post will summarize what I have found.

First of all, training a player to be a Catcher is really HARD. It took right at 150 games started before Santana had any rating at all at Catcher. After the 150th game, he went from no rating to a DEF of 10. Since then, his DEF has increased by only a bit over a point per year, as it now is about 14 (according to position calculator, it should end up as a 28 def, so he probably needs about ten more seasons before he will be there).

Second, when he is learning the position, he gets really tired. Until he got a rating at Catcher, he was resting almost every other game (he had a total of 91 games in his first full year). So it took him almost two years to get a rating at all.


So I have looked at total stats for each player over the five (actually 4 1/2 ) seasons (3304.2 IP at C for Santana and 3220 IP at C for Fisk) and this is what I have found:

1. Fielding percentage is very slightly better for Fisk (.987) than for Santana (.982) although RF is somewhat better (7.70 vs. 7.52)
2. ZR over those innings is -15.2 for Santana and -3.3 for Fisk. That sounds significant, but that is total over all five years (and each point is worth about a run) so it isn't dramatically different.
3. Santana has 27 Passed Balls compared to 13 for Fisk over that period. That comes to about 3 passed balls per year more.
4. Santana's CS % (27.5%) is actually quite a bit better than Fisk's (22.7%) over that period. Almost exactly the same number of baserunners tried to steal on Santana (404) as did on Fisk(401).
5. The real biggy is on Catcher ERA (CERA): Fisk's average CERA over the five years was 3.78; for Santana it was 4.28. That is a HUGE difference.

Anyway, anybody have any thoughts?

[Argonaut]

I'll keep my thoughts on this quiet in the hope that everyone continues to run Smoky Burgess.

What was most interesting to me is your CS% findings. 400 attempts is a pretty big sample, albeit just one sample. I wonder if OOTP doesn't take Catcher Arm into big enough consideration in its stat generation. Guys like Pudge should be throwing out 40% of runners and be noticeably better than weaker arms, all other things (pitcher's Hold rating, opposing Steal rating) held constant.

[allenciox]

By the way, one way of looking at this is the difference in defensive runs. These numbers are, of course, tentative, but it suggests something like:

1. Play in the field (as represented by ZR) cost about 12 defensive runs when Santana played as opposed to Fisk.
2. Passed Balls account for maybe another 5 defensive runs.
3. Figuring .055 run per inning based on Catcher ERA, Santana cost about 183 runs.
4. Santana may have saved 25 to 30 runs by throwing out more runners over that period.

So Santana cost, relative to Fisk, about 183+12+5-30=170 runs over the five years, or about 34 runs per year, equivalent to about 3.4 additional losses per year when Santana is catching. That is dramatic. But remember, this is one case, and it may not generalize. The actual CERAs might be much closer over an extended period of time and I will continue to monitor. If anybody else has tried this, let me know!

[Argonaut]

I just checked my Pudge and he has a 45% career CS%, so disregard my previous misgivings about Catcher Arm. But yeah, I don't think the true value of a defensive catcher shows up in ZR or WAR.

[chazzycat]

I have been running a Burgess / Posey tandem for a while now, currently on their 7th season together. Posey I've had even longer. Here are the stats, all perfect level:

Smoky Burgess (46 ability, 49 arm, 43 catcher rating):
- 5500 IP
- 4.42 CERA
- +2.2 ZR
- 36.4 CS%
- 42 PB (one every 14 games)

Buster Posey (86 ability, 70 arm, 90 catcher rating):
- 7700 IP
- 4.41 CERA
- +9 ZR
- 34% CS
- 49 PB (one every 17 games)

I really don't see much difference at all. My theory is that basically catcher defense only hurts you at the very low end..."average" defensive catchers like Burgess actually are fine for the most part.

I also think there is a lot of randomness in the arms. Posey's rating is 20 points better, but still lags behind Burgess in CS% even with seven seasons for the data to stabilize...I won't pay a premium for catcher arm, that's for sure.

I did also win a perfect championship like right after I switched from LaValliere to Burgess, so it's possible I'm biased

[Argonaut]

It seems like there's a lot of variability when it comes to defensive catching. And a good bat will be more consistent year-to-year. I tend to trust chazzycat's simulations he ran though, because the sample size was so large.

Of course you can just get a catcher that's good at everything like my man Buck.

[chazzycat]

Originally Posted by Argonaut

I tend to trust chazzycat's simulations he ran though...

See, there's your first mistake...