Pitcher Errors

Can anyone explain pitcher errors to me?

I understand what it means when you have a Leftfielder who has is listed as, say, 2(-2)e12. If he plays close to 150 or more games, he'll average 12 or so errors.

But what does it mean when a starting pitcher is listed as having 42 errors, or 11?
And what does it mean when a relief pitcher is listed as having 42 or 11 errors?

Thanks! Chompsky

If they are an e42, it means they will make lots of errors

e11 not so much

Seriously though, I have zero idea how they come up with the number! For example Bump Hadley made 4 errors in 47 chances in 1939 (.915) and he's an e37, however in 1928 he made 2 errors in 62 chances (.968) and his card is an e13 for that season. Seems like quite a jump for 2 more errors in 15 less chances.

Mike

I'm pretty sure it's based on batters faced, not defensive chances. Hadley faced almost 40% more hitters in 1928, it doesn't make up the gap between e37 and e13 completely, though, so maybe something else is normalized. Maybe throwing errors trying picking up a runner are included in wild pitches instead, and e-rating is restricted to fielding errors.

In terms of Strat chances, what you need to know is that the maximum of on-base allowed by pitchers' fielding is 2. You have the errors rating and you have the range rating. A rule of thumb to calculate the chances allowed is to take the excess of range rating over 1 and multiply that by 0.2 (easier, you double and move the point one spot) and add that by a third of the error rating, again moving the point one spot.

4e42...4 -1 = 3...3 * 0.2 = 0.6... 42/3 * 0.1= 1.4...0.6+1.4= 2 more chances, almost 2% if you look at the bar...

Strat formula is more complex, because errors and hits can add up simultaneously (single+error) and the range rating doesn't yield hits in a linear way (a 2 rating is closer to a 1 than to a 3). But the rule of thumb will give you a better estimate than not doing any calculation.

Correctly or incorrectly, I've always assumed that the E Rating is the number of errors the pitcher would make if he pitched every game. So staff full of e42 pitchers would reasonably be expected to make 42 errors over the course of an entire season.

If I'm wrong, please tell me.

That makes sense to me. You average the amount of errors of the starting staff, and that is the amount of errors you can expect from your starting pitching staff. Assuming they all pitch 32/33 games or 40/41 if they are * pitchers.

What about relief errors? If a relief pitcher is listed as having 40 errors, and the relief pitcher comes in every four games, should you expect 10 errors over the course of the year from that pitcher?

How is it modified based on different variables? One, the duration of the reliever? Two, the actual innings pitched by the reliever? Three, the number of times used over the course of the season?

Again, I could be totally off base, but I have always assumed that the rating is as if he pitched every inning in every game, because a relief pitcher who only pitched say 60 innings in real life could not possibly make 42 errors in those 60 innings. He probably only made a handful and that is "amortized" over the entire season.

Again, my belief, which could be totally wrong ...

Put another way ...

I believe that if your staff consists of all pitchers with e9 or lower ratings (just an example) I would reasonably expect 5-10 errors on the season from my staff.

But if I have a staff full of S8 and S9 types with e ratings in the 40's or 50's then I would expect to see 40 or more errors out of my staff over the course of the season.

The error rating is roughly based on how many errors a guy would make in 1400 innings (facing about 6000 batters), just like at other positions. So a guy that pitched 200 innings and faced 860 batters with one error would be an e7. They have wobbled on most things, including how pitcher errors are calculated, over the years, but in general his error rating will be close to (1400*E)/IP.

RE: error normalization...

Error ratings were normalized to rates of 1970. That is, players before 1970 generally saw their errors go down and after, the generally went up. I think the cards were normalized at least through ATG5. Not sure how SOM has managed this since taking over the online game from TSN.

...so Pitcher is kind of a strange position where error rates were actually lower in some years before 1970 (thus errors were normalized upwards). For the years 1928 to 1939, errors would be multiplied by these rates to normalize them to 1970:

1928 --> 1.07638565699022
1929 --> 1.07280906923173
1930 --> 1.08465788706491
1931 --> 1.08419649414341
1932 --> 1.06409550558158
1933 --> 1.04248700241136
1934 --> 1.04475439419264
1935 --> 1.04920328823925
1936 --> 1.04407302430481
1937 --> 1.03687578072391
1938 --> 1.04818625267322
1939 --> 1.07097654410086

For comparison, this is what normalization multiplier looks like for 2B from 1928 to 1939...

1928 --> 0.642516907867841
1929 --> 0.653815352053105
1930 --> 0.662048184401149
1931 --> 0.672984760198593
1932 --> 0.671622251841799
1933 --> 0.669984378582046
1934 --> 0.66979003154975
1935 --> 0.679203028817531
1936 --> 0.672922936928602
1937 --> 0.665580633946803
1938 --> 0.659122422096441
1939 --> 0.666290973618757