A couple thoughts to start…
1) On the whole I agree with Ted’s conversation. The game it attempting to “normalize” the ratings. Kind of.
2) It’s that “kind of” that can screw us up.
3) The main thing to know is that you are no longer seeing player’s actual ratings, but are instead seeing the game’s attempt to show a player’s ratings relative to each other.
4) My own view is that this approach is neither good nor bad, but merely different. IMHO, the game is not as good at this as some would like to say it is…but the concept of relative ratings is certainly solid, and if you squint just right it can feel more like real baseball.
All that said, things are not quite as rosy as they might oughta be. The game is not perfect in its application of the theory folks are tossing around. I should note here than I am not privy to the algorithms it uses, all can do is count ratings and show where it’s on and off—which is what I’ll do here.
Let’s start with
OVERALL RATINGS
All I’m doing here is counting the numbers of ratings of each 20-80 category, split by hitters and pitchers. Of particular note, I’m not yet cutting on specific position or, probably more important, starter and reliever. That should really be done, but I’m out of time today. Regardless, here are some histograms:
You can see they are not really true normal. Batters have this weird little double camel’s hump with a low at 45 and 50. Pitchers do, too, though it’s pushed down into the 45 category. There’s also some noise in the upper regions…and, of course, selection issues at the bottom of the scale. Baseball talent is not actually normally distributed in the composite.
Still, these are not horrible charts. I mean, I suggest we think of 55 as being the mean average required to really play a lot. 50 is okay. 40-45 make do, and 20-35 are your basic replacements and bit players you can squeeze extra value out of if you have sharp enough eyes.
Of course, these are just gross, overall ratings. They depend on OOTP getting that right, and to be honest, while I love their effort, I’ve generally thought the OOTP overall ratings were more visually pleasing than particularly bank-worthy.
So, let’s go to where the magic happens.
REGULAR COMPONENT RATINGS
First,let’s do something silly and just take every rating, regardless of whether it’s a CON or an EYE or a STU and throw them together into two charts—one for hitters and one for pitchers. It’s not useful at all for assessing BBA competitive advantage, but it’s a great exercise for seeing what the concept of relative ratings is trying to do. Here they are.
Ultimately thees look pretty nice. Hitters are centered on 6 and form a pretty nice looking normal curve. Pitchers are centered between 6 and 7 and also form something that looks normal. I note that this is why I pushed back on Ted’s commentary that “5” is the centerpoint that OOTP is shooting for. That’s not really right. If you’re trying to use just one number, it’s better to think of it as a “6”. However, that’s dangerous too, because (unfortunately for the theory) the average rating, even with relative ratings on, is not the same for each skill type. The distribution is more compressed than it is with absolute ratings, but it’s still not so cohesive that you can just wave your hand and say “5” (or “6”) is league average.
To see this, here are some messy histograms, showing the number of players in each component rating.
For hitters, Gap skews toward 7, CON is a 6, but EYE edges toward 5. AVK has a strong peak at 6, but skews toward 7. Power is on the edge of 6 and 7 as a center-point. The normality of the curves are pretty good, though—which is the major feature of relative ratings.
When we look at pitchers we see this even more readily. Stuff peaks at 7 and skews toward 8. Control peaks at 6 and is the most perfectly normal curve of them all, and MOV sits between 5 and 6. All of these ratings have longer tails on the lower rating scales.
SUMMARY
For my money, it’s helpful to realize that none of this actually changes the results on the field. The game still works with the raw data we’ve always seen before. To Ted’s point, this approach “cleans up” some of the confusion that can come from a league where ratings are moved off the norm by a ways…but as you can see, they can only go so far—you’ll still benefit to a degree based on how you read the ratings.
I hope, however, that some of this gives you a better idea of how you want to read the ratings, because part of the charm of the game is that the way you read the ratings will be different than how I read them (Gap, for example, is valuable to me, not so valuable to Ted…that’s not going to change). Understanding that the “average pitcher” is not 5/5/5 or 6/6/6, but is probably 7/5/6 now, and knowing that the ratings will still shift a little over time is still going to be useful.
