What does this exactly track?

[tigers527]

Ok bear with me I was trying to measure a point that maybe teams with more errors generate more overall runs vs earned runs. I thought the easiest way to do this would be to ER/R and then look at a few other numbers.

percentage of Runs given up that are Earned, ranks in league
1 NYY 94.04% 8th ERA, 1st in errors, 3rd in FP (fielding %)
2 CLE 93.40% 9th ERA, 6th in errors, 5th in FP
3 BAL 93.30% 13th ERA, 8th in errors, 7th in FP
4 SEA 93.10% 11th ERA, 10th in errors, 10th in FP
5 BOS 92.72% 5th ERA, 2nd in errors, 1st in FP
6 TBR 92.46% 2nd ERA, 5th in errors, 6th in FP
7 OAK 92.34% 4th ERA, 9th in errors, 9th in FP
8 LAA 92.15% 3rd ERA, 4th in errors, 4th in FP
9 KCR 92.10% 10th ERA, 7th in errors, 8th in FP
10 TOR 91.83% 1st ERA, 3rd in errors, 2nd in FP
11 MIN 91.72% 7th ERA, 11th in errors, 11th in FP
12 DET 91.25% 12th ERA, 13th in errors, 13th in FP
13 CWS 89.53% 6th ERA, 12th in errors, 12th in FP
14 TEX 88.58% 14th ERA, 14th in errors, 14th in FP

The only thing I think I tracked for sure is that TEX belongs on the bottom, and they are. May the entire bottom of the list is correct. I think I might only be tracking which team has the nicest score keeper, I think they might reside in the AL east as 4 of the 5 east teams are in the top six. If your gonna score against the Yankees its most likely to be earned.

Anyone have an opinion on these numbers?

BTW numbers collected a few days ago so with tiny variation I would imagine they are similar. And ranking in errors is fewest to most (1st=lowest, 14th=most)

[terpsfan101]

Tigers527,

You could very well be "tracking which team has the nicest scorekeeper." If you are looking for a way to measure team defense, I suggest you take a look at Defensive Efficiency Ratio (DER). DER tells you the percentage of batted balls that the defense turned into outs.

DER = (BFP - H - K - BB - HBP - 0.6*Errors) / (BFP - HR - K - BB - HBP).

[tigers527]

Tigers527,

You could very well be "tracking which team has the nicest scorekeeper." If you are looking for a way to measure team defense, I suggest you take a look at Defensive Efficiency Ratio (DER). DER tells you the percentage of batted balls that the defense turned into outs.

DER = (BFP - H - K - BB - HBP - 0.6*Errors) / (BFP - HR - K - BB - HBP).
I think what I wanted to measure is that more errors = more unearned runs, what I wound up with does not necessarily support that thought. Although the Yankees with the fewest errors give up the smallest % of unearned runs, and Texas with the most errors gives up the largest % of unearned runs. So those two results jive, it is 2-10 that kinda make me scratch my head and make me wonder what I am looking at.

I guess maybe the sample sizes are too small?

[SABR Matt]

Random chance.

The problem is that once an error has occurred, it is relatively random how many runs follow it. Sometimes an error happens with two outs and the team goes on to score 11 runs in the innings and all of them are unearned even though there was only one error...sometimes an error produces no runs at all.

[tigers527]

Random chance.

The problem is that once an error has occurred, it is relatively random how many runs follow it. Sometimes an error happens with two outs and the team goes on to score 11 runs in the innings and all of them are unearned even though there was only one error...sometimes an error produces no runs at all.
Even in the unlikely 11 R scenario that would be less than 1.5% (enough to drop a team 6ish places) deviation for a 600 runs allowed team.

But you are right errors do not automatically lead to unearned R, but they can.

[SABR Matt]

That's just one inning though...some teams just get unlucky in how the events cluster following their errors...it can happen any number of times in a season.

[SABR Matt]

My point is...if you were to take each error that occurred in a game and count the number of unearned runs that followed it in that half inning, you'd get a distribution with a mean of (say) 0.4 runs but a wide range of actual possibilities...it would probably be an exponential decay function with peak at zero runs following and a long low probability tail after 2 runs.

[Colin Wyers]

Random chance.

The problem is that once an error has occurred, it is relatively random how many runs follow it. Sometimes an error happens with two outs and the team goes on to score 11 runs in the innings and all of them are unearned even though there was only one error...sometimes an error produces no runs at all.

It's not really random at all, although as with everything there are sampling issues. For the most part, a pitcher will give up unearned runs in proportion to how many earned runs they give up - a pitcher with a lower ERA will (usually) have a lower UERA. The ability to prevent runs after an error is the ability to prevent runs.

[SABR Matt]

Not entirely true Colin...if that were entirely true, then Unearned runs / total runs would be a constant...look at the first chart in this thread...it clear is not a constant.

I'm sure some component of UER prevention is tracked by ER prevention, but there is GOING to be random variability.