In the past few days, reader Bill Ferguson and I have been discussing the finer aspects of APBA’s methods of grading pitchers via email. It’s been very interesting and we thought we would share some snippets of our conversation to open it up to the larger community since I know there are others who share similar feelings when it comes to pitchers’ grades.
Bill initially wrote me:
I think [APBA] has done a poor job of pitcher ratings relative to the batter cards. You’re basically, giving one rating to the pitchers based upon ERA whereas it should be based upon hits/ip.
ERA is a product of fielding ability of the players behind the pitcher – double play frequency, arms in the outfield, etc, the frequency of walks, the likelihood that a reliever will close down the offense stranding runners left on base by the pitcher or letting them score.
So what I do is assign a pitcher a rating that is purely a reflection of his hits/ip and that rating is used whenever starts and inning until he gives up a hit, a walk or an hit batter. Then he uses his APBA rating. So I get a mixture which should reflect the best of both worlds. For this purpose I assume an average pitcher will get an 8 rating.
My response:
I don’t necessarily disagree with you and I know a few people who feel strongly that APBA should base grades on the pitchers’ actual performance such as hits and walks given up and not on stats which are the possible result of his performance (like ERA) or stats that are dependent on external factors (like wins).
That said, ERA is simple and easy for everyone to understand. Not every baseball fan wants to bother with h/ip. I know it sounds simple but quite honestly, if you poll your casual baseball fan, most probably won’t know what is a good WHIP and was isn’t.
Again, I’m not disagreeing with you. I’m just looking at it from APBA’s point of view.
But that is the beauty of APBA. If we don’t like a system, we can modify it. What is the system you use? How do you rate the pitchers?
Bill’s response:
I understand (and somewhat agree) with your point about simplicity. However, for someone to play the master game of APBA, he’d have to be more of a student of the game. He’d have to be a detailed person. It’s that point of view that I base my idea. Put in the simplest of terms, Boggs of StL has a 17 rating. Motte has a 14 rating. If you look at their relative hits/ip you’ll see Motte is significantly better than Boggs. This is just one example. I could name several (if not hundreds) of examples. The only reason Boggs has a higher rating is ERA. Motte’s ERA was so great because he had a tendency to give up the home run ball. That justifies is L rating. But the L rating itself and the fact that Motte had nobody to clean up after him caused his ERA to be higher than Boggs. Was Boggs that much a better pitcher? I doubt. If (and when) Boggs got into trouble he more or less got bailed out by Freeman or Rzepcysnski.
You might say why all the fuss? It’s just a game. Well, here’s the fuss. Do you then use Boggs rather than Motte to be the closer? Is Wainwright truly just a low C grade pitcher whereas Lynn is a solid B? Lynn’s hits/ip were 8.6; Wainwright 8.9. This hardly justifies the difference in rating from 6 to 11. That’s a little more than 30 points in batting average.
The real reason for my concern is there is just a sense of despair when one team has nothing but D rated starters (for the most part) and plethora of B or A rated relievers. If the relievers were all that good they’d be making more money and learning to start rather than relieve. The problem is many of the starters have their ERAs inflated because a reliever didn’t shut down the offense when the starter left men on base.
Bottom line is this. Variety. Batters are rated for walks, Ks, homers, doubles, singles, DPs, etc etc. Pitchers have two really useful rating – the basic grade and the Z rating. X,Y,Ks are now become valuable to eliminate sacrifice flies but, even still, that only 3 ratings. My second grade for all pitchers adds some variety and closes the gap between the undergraded pitchers and the overgraded pitchers, which is a lot closer (except for Verlander and people like that) to real life.
Here is the chart Bill uses:
| Hits/9IP | MG Grade |
| 5.5 | 17-19 |
| 6.5-7.5 | 14-16 |
| 7.5-8.5 | 11-13 |
| 8.5-9.5 | 6-10 |
| 9.5-10.5 | 4-5 |
| 10.5+ | 1-2 |
Bill’s notes on his chart:
The reason for the "drop" between 11-13 and 6-10 is the fact that on Bases Empty on the APBA board there’s a 60 point drop in batting average as we exclude #9 from being a hit. Each batter has two 9s so that about 60 points on the BA using about 30 to 31 ABs.
As I noted above, I do sympathize with Bill and know he’s not the only APBA fan who feels that pitchers’ grades should not be based on ERA or wins. We’ve broached this topic before, I believe. The ERA statistic while useful, is not truly reflective of a pitcher’s effectiveness but rather one of the symptoms of his performance. Unfortunately, other factors play into it as well like Bill says. Wins, are even more dependent on other factors.
Like I told Bill, I can’t fault APBA for using ERA as the grading method. Like it or not, it’s the standard MLB and the sports media uses to judge the pitchers (though they are catching up in recent years). If APBA wants to appeal to a wide audience (i.e. the casual fan or the young fans), they must use the current standards in place.
This may not be the place to bring this up but APBA’s method of assigning control letters (Zs, Ws and ZZs) by using walks per batters faced instead of per 9 innings pitched seems quite advanced compared to the grading system… sometimes to my frustration :)
And again, this is one the most appreciable aspects of the game of APBA. It is very customizable and open to modifications. Don’t think that fact is beyond CEO John Herson. In fact, he probably counts on it. Not only does he appreciate the innovation within the APBA community, my guess it that it helps him in future development of the game.
Thanks to Bill for spurring this discussion and sharing his chart. I’d like to hear from others who either agree or think that the grading system is fine the way it is. Me? I don’t really think it’s ‘broken’ but but am always looking for ways to improve it.
The goal in all cases should be to accurately replicate a pitcher’s ERA. That’s why there’s kind of a problem in giving a Basic rating based purely on H/9. Because it wouldn’t take into account HR/9.
For example, you might have two pitchers who each allowed 8.9 K/9 but one of them allowed 1.8 HR/9 and the other was only 0.5 HR/9. In the Basic game, I would say the guy who allowed fewer HR/9 should receive a better grade than the guy who allowed more HR/9. Even though there H/9 were identical, you need to account for the HR/9 which the Basic game doesn’t do.
Now in the Advanced game, they should both have the same MG rating.
Note that I think that in the Basic game, for example, those guys might be a C and B pitcher, respectively. Although it’s possible that their MG ratings both are something like a 12.
To bring up a rival game company, for example… Strat-o-Matic has a Basic and Advanced version of play. In the Basic version, all fielders receive a rating of between 1 and 5 – 1 is the best, 5 is the worst. In the Advanced version, they receive a 1-5 rating to reflect their range and an ADDITIONAL number that reflects how often they commit errors. What this means is that there is often a player who, for example, is a 2 in the Basic version but a 3 in the Advanced. For example, a guy with limited range in the outfield but who never makes an error (like a Ryan Braun for example) might be a 3 in the Advanced version but a 2 in the Basic version.
I think APBA’s pitching grades should probably work similarly.
In the Basic version of APBA, I think the card creators have to make a choice between getting accurate H/9 results or accurate ERA results. There’s no way for them to do both aside from cases in which a pitcher’s HR/9 is right around the league average anyhow. My preference is that they get their ERA results accurate rather than the H/9. If I want both, I’ll have to play the Master Game.
As a side note, basing all of these things on some number per 9 innings is problematic – it should be based on H/BF, HR/BF, K/BF, etc. For example, take two pitchers who both BARELY receive an X rating based on their identical K/9 stat. One pitcher is a D and the other is an A. If each throws 200 innings, the D pitcher will have to face more batters than the A pitcher because he puts more men on base. The yield? The D pitcher will therefore end up with more strikeouts in 200 innings than the A pitcher when he should have had the exact same amount.
This isn’t to dis on APBA, by the way. I do about 40% Strat, 30% APBA, and 30% Diamond Mind and could probably write 100 pages on different elements of each game that are stronger than the others. I’ve played thousands and thousands of games with each and can honestly say I believe that there is no perfect sim.
The goal in all cases is to replicate ERA? ERA is a severely flawed stat. With 2 outs, Smith makes an error. The pitcher then gives up 500 consecutive home runs. His ERA is zero. Come on.
When I wrote “8.9 K/9” above it should have read “8.9 H/9”. Sorry about that.
One other point against using ERA as the basis is relievers. Guys who sometimes enter a game with 1 or 2 outs already have artificially lower ERA. Starters don’t that advantagr. Every inning they appear in started with 0 outs.
So a reliever who had 11.5 h/9 can still have a lower ERA than a starter with 9.5 H/9.
Since I draft players for my own solitaire 8 team league instead of doing replays it would be better for drafting if they carded players for individual rather then team performance but at the same time I don’t want to see the basic game become more complex. That’s why I’m not interested in the newer games out there. So to sum it up maybe its best to just leave things as they are. Just my opinion.
Example? Not sure what you mean by carding them based on individual rather than team? You mean a pitcher with barely A numbers should stay an A rather than becoming a B because he was surrounded by Gold Glovers?
Great discussion, guys!
I think that Bill’s point about variety is a key one. Back in ancient times I looked at career statistics for players that I’d owned for a good number of years (in a league that used the basic game) and found that hitters tended to arrive at career stats that were eerily close to their actuals, while this seldom happened with pitchers. My guys Ted Simmons and Keith Hernandez were right on the money, which should reveal my Cardinals proclivities and my advanced age! I think a big part of it is that the hitters’ cards drive the game and that they have many ratings and 600 tries in a season and 10,000 or more tries in a long career. Pitchers had basically two ratings, and one drubbing could inflate the ERA for the year to a point it would never recover. That and you could get an A grade for 20 wins and a 4+ ERA.
Another (master game) league I was in for a number of years tried to remedy this with its own system. The initial letter grade was ERA-based and according to the commonly accepted formula of the day. BUT the league devised (or copied from someone, I can’t remember) a system that attempted to refine walks and home runs based on numbers allowed per inning. Each pitcher got a rating akin to the master game steal number. A pitcher rated G24 would get a roll on every home run result; a roll of 24 or lower would turn the HR into a double. The flip worked as well; an L24 pitcher would roll on all doubles, and a low roll would turn the 2b into a HR. Same concept for walks. Finally, these HR and BB factors were used to adjust the base ERA that determined the final numerical pitching grade. I don’t recall the amounts exactly, but if a pitcher were rated G36, expecting to turn half of home runs into doubles, that would lower his grade by two or three points.
This wigged out guys who didn’t get the concept. I recall one example, but don’t remember who the pitchers were–but they had about the same ERA. One of them gave up few walks or home runs, while the other was neutral. The former ended up with a grade maybe 4-5 points lower. The guy who owned the pitcher with the lower grade actually quit the league over it; he should have just traded for the guy with the higher grade!
Anyway, I think it did sort of accomplish the goal of making pitching results more realistic. But it also hurt the playability of the game. With so many extra dice rolls, the games took longer to play, especially as offense exploded in the 90s.
I believe that Diamond Mind, the computer version, is a far superior simulation. It does a great job of taking in so many different factors. Does anyone remember the board-game version, Pursue the Pennant? It may have had a different name at first. It was virtually unplayable as a board game. APBA is more fun, but the more you tinker, playability suffers.
I enjoy these types of exchanges as they demonstrate an interest in both APBA and the game of baseball.
My three thoughts
1. APBA’s Basic game is robust enough that it can be customize it in a manner that reflects the interests of a particular individual and/or members of a league. These customizations are made easier by the existence of a number of online resources and our ability to quickly share them.
2. The Basic game is intended to reach a range of customers. To stay solvent the company needs to attract the next generation of customers, which means keeping things simple enough that a seven year old can take the game out of the box and begin having fun.
3. The intent of the Basic game (Master game too) is to replay a season. Because of that intent a pitcher’s grade not only reflects their ability but to a certain extent their team’s ability as well.
John Williams
While ERA is obviously influenced by many things, to pretend it’s not at least some indicator of quality is silly. While there’s a difference between pitching in Oakland and pitching in Wrigley Field, if ERA is really “a product of fielding ability,” DPs, arms and bullpens, how come different starters on the same team (who pitch in front of the same fielders and in the same parks) often have very different ERAs?
You act as if the pitcher is just an innocent bystander and his ERA is only determined by those outside influences.
Hi Kenn,
I think John’s point is somewhat valid. ERA is one step removed from the stats of what pitchers actually do.
Case in point, I follow college baseball, in particular the U of Illinois. While they’re on the road, I depend on boxscores and play-by-play accounts. The other day, our reliever pitched 1 1/3 innings with no ER. Sounds great, right? Well, he gave up two hits that drove home a couple runs that weren’t charged to him.
Long story short: I’ve learned some time ago when recapping away games not to rely on stats and box scores and instead comb through the play-by-play. :)
I realize that’s a one game example but you get my point.
Basically, stats are just a tool for us to describe our reality, in this case baseball. I do agree that for the most part, ERA is effective (though more so for starting pitchers and not so much for relievers).
Bill’s method is an interesting idea though.
The point I am attempting to make is that APBA like the other makers seeks to produce a game that can be used to replay a season.
The set of cards created for each team are intended to recreate that team’s performance.
The season replays reported in the old APBA Journal and now in the Delphi’s “APBA – Between the Lines” suggest that APBA’s approach seems to work.
The teams tend to perform in a similar manner.
Hi guys,
I have enjoyed the discussion because I have thought abt some of those same things in my own solo play. I’m wondering how much fatigue factors would play in keeping numbers close to the reality we want them to reflect. I’ve tried to bring fatigue into my play & it does make gms longer but it does make me think in terms of how long starters go & what relievers to bring in. I like tht APBA mages a product tht we can modify it however way tht works for a particular league or solo play.
After all we all play for the love of the game, the relationships we’ve made, the compiling of stats, ect..
A labor of love:)
Shawn Kaufman
Should read APBA makes a product tht we can modify it however we need to:)
Fatigue & pitch counts go into the equation as well. I am thinking abt David Price vs. Justin Verlander for the Cy Young. I’m not saying Price didn’t have a great season & didn’t deserve to win but Verlander pitched in war more innings than Price, so what happened in gms after Price left the game….my point is Verlander was in gms longer & more opportunities to give up runs & hits. Price pitched less but certainly dominated while he was in there bt who was more valuable to their team…just some thoughts:)
Shawn Kaufmann
First let me say that there never will be a perfect system. Undoubtedly the strength of the defense behind the pitcher impacts ERA. Of course it also impacts hits per inning. One only had to watch Ozzie Smith during the 1980s when he probably prevented a hit a game, or at least it seemed like he had one unbelievable play every day. The W, Z and ZZ seem adequate to address walks. So the issue is hits and I have to agree that grading should be tied to hits more so than ERA. I personally focus on individual achievement rather than team, although I am strictly a replayer. I prefer the APBA system to others because there is more differentiation between individual pitchers in the master game numbering system. I also agree with the comment that we are free to modify the # as we see fit. That is why I was so glad that 5.75 allows complete editing.
I’ve gone back and forth over the years about APBA and what to do with some of its perceived shortcomings. I even had the master game at one point. But then I realized all this time spent deciding what tweaks I should use takes away from the time I should be playing the game. I also don’t have the time and, honestly, interest to regrade an entire season.
I also found another game, Replay (tried Strat, didn’t like it) that gave me more options especially with pitching and fielding. I play both.
But what it comes down to is APBA basic baseball is just that, basic. It will replicate pretty well a teams performance over a season. But it won’t be that detailed in many respects. And that is fine with me. It’s a FUN and fast game to play and stats come out pretty good, even for pitchers.