Joe is the founder of Project COBB, under which he serves as Stats and Research Manager for the GB National Team and chairs the British Baseball Hall of Fame. Joe's writing has been published in book form by Fineleaf.
Please use the following buttons below to share the post that you are reading with the popular aggregators:
The weekly You Are the Scorer post will be used as a thread for scoring queries for the rest of the year, so please leave a reply below if you have a question or comment.
To hunt through the first 50 You Are the Scorer questions, click here.
Tags: Keeping score
Hi Joe. This scenario came up in an MLB game I was watching back over the weekend.
It was the typical double play set-up at second. The second baseman fielded the grounder and chucked it to the shortstop, who got the force but then dropped the ball on the transfer. The runner was called out at second and the batter-runner was safe at first, with no attempt being made to put him out.
Now, I’m fine with the 4-6 put-out, but the commentators never raised the possibility of an error. It looked to me that the shortstop had time to field the ball and complete the double play and he didn’t because he dropped the ball. Can that be called an error or is it a case where you can’t assume the double play would have been completed?
Alright, I have a question about fielding stats – which I feel are severely underdeveloped – and wanted to know your thoughts.
Here’s the scenario. I will give you a series of plays by an imaginary left fielder in a busy 9 inning game, and I would be interested to know what stats (fielding percentage, range factor etc) you could draw from it.
1st inning: routine fly ball caught
2nd inning: routine fly ball drops out of glove
3rd inning: cuts off routine single and throws the ball safely to 2nd
4th inning: fantastic diving catch – throws to 2nd to catch the runner for leaving early on appeal play
5th inning: diving stop brilliantly holds a sure triple to single and saves a run. Throw in to the infield is too long and runner advances to 2nd.
6th inning: misplays a routine grounder: runner advances to 3rd base
7th inning: loses a fly ball in the lights and fields it on the hop – runner is given a single
8th inning: runs underneath a linedrive and gets nowhere near it. Other outfielders are of the opinion that it was easily catchable.
9th inning: cuts off a ball in the gap, makes a great throw to 3rd. The runner is forced to turn and head back to 2nd.
Thanks for another great question.
While there’s no way of changing the established staple baseball phrases now, the statement “you can’t assume a double-play” would be more helpful – although admittedly a lot more clunky – if it was instead something like “you can’t assume a double-play unless a good throw for the second put-out is dropped by the fielder at the bag.” So in your case, the fielder who makes the misplay is deemed to have done enough on the play to be exempt from an error. The only time we would award an error on the attempt to make the second out on this play is if the throw was good and in time, but the first baseman dropped it.
Thanks for the question. I agree with your sentiments about fielding stats, although progress is certainly being made in the Major League with stats like UZR (http://www.fangraphs.com/blogs/index.php/uzr-on-fangraphs/).
The range factor I’ve introduced to British stats is cruder than Jethro, but it is still better than just having fielding average. It is simply worked out as put-outs plus assists all times 9 and then all divided by the number of innings played. So it bascially gives chances converted per game. In contrast, fielding average is put-outs plus assists divided by put-outs plus assists plus errors. So that can be seen as proportion of chances converted out of all chances.
So let’s now run through your excellent example.
1st inning: 1 put-out
2nd inning: 1 error
3rd inning: no stats recorded
4th inning: 1 put-out and 1 assist (as long as the appeal is not initiated by the pitcher)
5th inning: 1 error (a harsh statistical reflection on a positive outcome overall
6th inning: 1 error
7th inning: no stats recorded
8th inning: 1 error (unless the scorer was feeling kind)
9th inning: no stats recorded
Let’s first total the stats up over 9 innings: 2 put-outs, 1 assist, and 4 errors. This gives a range factor of 9 * (2 + 1) / 9, which equals 3.00. The fielding average is (2 + 1) / (2 + 1 + 4) = 0.429.
The fielder’s strengths are not really picked up in either stat, so the weaknesses (which are clearly pretty major) show through prominently. I would suggest that this fielder needs to consider playing in a league with a DL spot.
One thing we used last season was to work out an “alternate” fielding percentage, which calculated errors as a percentage of the total “chances to make an error” – ie in my opinion, misfielding 50% of balls to the outfield should result in a .500 fielding percentage (but of course it doesn’t) because unless an out is made it doesn’t count towards “total chances”. We felt that this gave a better indicator of fielding ability than the orthodox fielding percentage. you can also calculate separate groundball handling, flyball catching, and throwing percentages. This demonstrates what you need to work on with the fielders in question.
Other ideas we used were a +/- system with a + being awarded for every outstanding play, and a – being awarded for a mistake (even if it wasn’t necessarily an error). Obviously this is rather subjective. It works especially well for base running stats – going from 1st to 3rd on a single may get you a +, failing to tag and go home on a deep flyball may well earn a -.
Position adjusted range factors are reasonable indicators – we also like zone ratings, which analyses the percentage of hit balls which go through a fielders predefined “area” that he makes a play on successfully, semi-successfully, or unsuccessfully.
All this is quite complicated: I had to design an entirely different scoring book to make sure everything was recorded accurately!
I’m fascinated by your approach – and your amended scoresheets.
I must admit that I’ve considered such a system but in the end have decided that the samples we deal with in each season of baseball in Britain are so small that noise will almost inevitably drown out any real information. Therefore, the fielding awards we assign in the National Baseball League are based in part on subjective ratings of range/arm, rather than any advanced statistics. I feel that this is the safest option. That’s not saying that there it is not still interesting to play around with advanced fielding stats in our leagues.
Thanks for your thoughts,
Mail (will not be published) (Required)
© 2014 BaseballGB. All Rights Reserved.
Powered by WordPress. Designed by