The stats projections for the 2023 season are now up. Actually they first went up around Christmas, but I’ve been doing an unusual amount of tinkering this time around. Updates should happen at least a couple of times a week between now and the start of the season – which kicks off on March 30 this year.

1. Rules changes for 2023.

Trying to figure out how the reduced shifting and the new pickoff attempt rules will affect the statistics was definitely an exercise. A lot of it, admittedly, is guesswork, but we do have the results from the minor leagues to guide us. The gist of it is I raised stolen base attempts for all players by 25% (which obviously makes a bigger difference for guys who steal 20 bases than guys who steal 2), while decreasing caught stealing rates. Left-handed hitters received improved single rates, enough to raise their batting averages by about 10 points. Switch-hitters got a bout a 7-point raise. Righties get nothing. And similar adjustments were made for pitchers, with right-handed pitchers (who face more lefties) getting a slightly larger downgrade on their BABIP.

2. Pitcher adjustments

Every year, I make adjustments made to the way I translate between the player statistics and what I expect in the next season. The changes in the pitcher numbers are somewhat larger this year. I made a mistake last year, which did not become fully obvious to me until later, in that I way over-did the regression to mean aspect. All pitchers, good, bad, or otherwise, got squeezed into a tight bin, with not nearly enough variation. I tried to avoid that trap this year.

3. Stronger adjustments for foreign leagues

There are some oddities in projecting statistics from Japan, and I took a deep dive into them this year after a disappointing performance from some more disappointments. I’ve always calculated league strength by using the results of all players who played in those leagues. I had recognized a problem with that before – that the translations between AAA and Japan gave very different results than those between the majors and Japan – but extended it this year to noting where players are from. The result of that is Japanese hitters losing about 20 points of EqA in their translation. I will try to drw that out in another post.

4. Fielding changes

A friend of mine pointed out that the fielding numbers in the translation seemed off. This was something that I think I’ve known for a while. I had built in a sort of universal fielding adjustment to handle “defense” without having to track the position closely, and it just didn’t work. So I totally overhauled the way that worked, which should lead to more consistent values.

5. New schedules

The first set of numbers I put out still had the 2022 schedule built into it. Most years that is no big deal, but this year they are changing to a more balanced schedule, with less concentration on your own division. That was good for teams in the strong Eastern divisions (the Nationals gained 0.77 wins when I changed the schedules, most of any team; the Marlins, Red Sox, and Orioles all gained over 0.5 wins). And it was bad for the AL Central (the Guardians and White Sox were the two biggest losers, dropping by 0.85 and 0.69 wins). Those changes in team totals will be accompanied by changes in individual numbers, but it isn’t large enough to be a real issue.

6. org Files with Roster Status

One of my personal favorite tools while gearing up for the season are the org dt pages. Each team is broken down by position (with pitchers grouped into Starters, Relievers, and Swing), and each player is ranked in order of their projected Major League playing time. I find it a good way to see a team’s entire depth chart – including their coming league players. And I now have it cross-referencing with MLB roster status, so you can see who is on the 40-man roster (“A”, until we get into the season), who is an NRI (“n), and who is injured (“D”, when we get that far). Here are Brewer shortstops:

    SS
  Last         First       Team Lg  Age  PA AB   R    H  2B  3B  HR RBI  BB  SO  SB  CS   BA    OBP   SLG  EqBA EqOBP EqSLG  EqA   VORP  WARP Defense FRAA  MJ BRK IMP CLP ATT DRP UPS 
A Adames       Willy        MIL NL  27  597 543  81 147  28   2  28  89  53 170   7   4 0.271 0.337 0.484 0.272 0.340 0.488 0.284  43.1   5.2 140-SS   4    94  10  42  35  18   8 133 
n Alvarez      Eddy         MIL NL  33  433 386  50  99  16   2  10  39  35 117   7   2 0.256 0.337 0.386 0.266 0.345 0.394 0.264  20.6   2.2 100-SS  -1    61  16  36  56  26   8  20 
A Turang       Brice        MIL NL  23  625 576  69 152  25   3  10  60  49 143  22   3 0.264 0.322 0.370 0.264 0.324 0.363 0.256  25.1   3.4 147-SS   5    40  27  57  30   8   1  54 
n Monasterio   Andruw       MIL NL  26  437 398  53  97  18   2   9  43  35 116   9   2 0.244 0.311 0.367 0.258 0.326 0.375 0.249  14.1   1.3 102-SS  -3    34  23  56  48  25   6  33 
  Devanney     Cam          MIL NL  26  587 539  59 117  26   2  17  59  41 165   5   2 0.217 0.281 0.367 0.231 0.294 0.376 0.233  10.1   0.9 137-SS  -2    24  31  52  41  24   5  32 
  Brown        Eric         MIL NL  22  305 278  33  58  14   2   6  23  22  81  19   2 0.209 0.279 0.338 0.226 0.295 0.343 0.237   6.4   0.8  71-SS   1    15  23  40  40   1   0  50 
  Zamora       Freddy       MIL NL  24  231 218  20  46   6   1   3  13  10  63   6   1 0.211 0.255 0.289 0.211 0.254 0.284 0.202  -2.7  -0.4  54-SS  -1     5  64  74  27  20   1   5 
  Garcia       Eduardo      MIL NL  20  541 516  52 110  22   2  13  53  20 217  12   1 0.213 0.250 0.339 0.227 0.263 0.350 0.214  -0.5  -0.3 127-SS  -2     2  43  65  30   8   4  15 
  Murray       Ethan        MIL NL  23  427 393  38  79  16   2   8  35  31 137   9   2 0.201 0.265 0.313 0.215 0.279 0.315 0.214  -0.7   0.5 100-SS   5     2  28  51  37  13   2  11 
  Barrios      Gregory      MIL NL  19  363 341  36  78  13   2   1  28  19  82  13   1 0.229 0.275 0.287 0.233 0.279 0.289 0.214  -0.4   0.9  85-SS   8     0  61  76  22   6   1  13 
  Guilarte     Daniel       MIL NL  19  266 246  20  54   9   1   1  21  20  79   9   2 0.220 0.278 0.276 0.228 0.288 0.281 0.211  -1.0  -0.3  63-SS  -2     0  49  62  27   3   0  13 
 

One Response to Statistics projections for 2023

  1. clayd says:

    Made a further update to revise the Upside calculations.

Leave a Reply

Your email address will not be published. Required fields are marked *

Set your Twitter account name in your settings to use the TwitterBar Section.