Sunday, May 01, 2005

FAQ on ERV Scoring

[Ed. Note: Here's is the first rough draft of the ERV Scoring FAQ. More will be added as we go along, but this hits the very high points]

Version 1.0, Date 5/1/2005

What does ERV stand for?

ERV stands for Expected Run Value. There is a thing called an ERV matrix, which is a 3x8 chart (like the one in this article) that tells you how many runs the average team scores at all of the "states" of an inning of a baseball game. A "state" is simply the runners on base and outs in the inning, e.g. none on, 2 outs or bases loaded, 1 out. The ERV is compiled by looking at every game and every situation and averaging the actual runs scored after that situation. For example, in the 2004 MLB season, with bases loaded and no outs, the average team scored 2.25 runs in the remainder of the inning. So, the next time your team has bases loaded an no outs, if they don't score 2 or more runs that inning, you should be disappointed.

What is "ERV Scoring"?

It's a way of using the ERV matrix while you keep score in a baseball game. For every play in the game, you can compare the state of the game before and after the play, and, by comparing the values from the ERV matrix, you can assign a value to the play. For example, with none on and no outs, the average team in 2004 scored 0.54 runs. With no outs and a runner on first, the average teams scores 0.93 runs. So, if the leadoff man walks, you can give that walk a value of 0.39 runs, the difference between the "none on, no out" state and "runner on first, no out" state. We call that 0.39 the Run Value, or RV, of the play. You can credit the batter with 0.39 RV, and debit the pitcher with a -0.39 RV. For a complete description of ERV Scoring, see the Guide to ERV Scoring (coming soon).

So, what is the difference between ERV and RV?

ERV is Expected Run Value, or the runs the entire team is expected to score in the inning. RV is the value you assign to any particular play, and what you total up for batters and pitchers after the game is over.

Why would anyone want to do this?

Because it's fun, that's why! If you like keeping score at games, this is a lot more interesting than keeping track of pitches, for example. It is relatively simple to do, and it gives you some more clear insight into which plays are really valuable and important during the game, and which are not. For example, when a pitcher throws the ball away on a sure double play ground ball with one out, ERV Scoring tells you it cost the team 1.4 runs. Also, the total RV for a player tells you precisely, in a single number equivalent to runs, how much a player contributed to the team's success (or failure). For example, if a player's RV is negative 2.5 for the game, and the team lost by 2 runs, it is a safe conclusion that he lost the game for them.

But how can anyone get a negative RV? What is a negative run?

Remember, ERV is based on the league average -- the average team scores 0.53 runs per inning. So, if the team doesn't score any runs in an inning, they have not met the average expectation, so we say they scored -0.53 runs. Also, for batters, if they make an out, they've put their team in a worse position for scoring runs, so it makes sense that the batter gets a negative. ERV Scoring is analogous to the plus/minus stat in hockey, but it is much more precise, and if a player has a negative RV, he is clearly not meeting average expectations.

You said an RV can be assigned to any play, does that mean fielding and baserunning, too?

Yes, that is one of the most interesting features of ERV Scoring -- you can assign values to errors, great plays, stolen bases, taking the extra base, etc., simply by comparing the states before and after. For example, to isolate the value of an error, simply compare the state of the game had the play been made, with the state that actually ended up. For example, let's say the leadoff man grounds to short, but the fielder bobbles the ball for an error. Had the play been made, the batting team's ERV would have been 0.28 (None on, 1 out). However, the error means it is now 0.93 (Runner on first, no outs). Therefore, the error cost the team in the field 0.65 runs (0.28 minus 0.93). You can assign that value to the fielder, and give the batter the RV he deserved (-0.25, or 0.53 minus 0.28) because the play should have been made. Also, that RV does not go on the pitcher's record because it was not his fault that the play was made -- he gets credit for coaxing a ground out.

One way to think of ERV Scoring is a system to break up the runs scored in a game into little pieces that can be distributed among all the players (pitcher, runner, fielder, batter) appropriately, so we all have a better understanding of who is responsible for victory and defeat.


Is ERV Scoring a "zero-sum" system?

Yes! The RV for a team in each inning should add up to a value that keys off of the 0.53 figure. For example, if the team scores 1 run, the RV should add up to 0.47, which is 1 minus 0.53. That team scored 0.47 runs above the average. For 2 runs, the RV should add up to 1.47, or 2.00 minus 0.53. This helps you keep track that your individual player RVs are correct. Also, the RV assigned to the batters should equal the opposite of the RVs assigned to the opposing pitchers. The Guide to RV Scoring explains this in more detail.

12 Comments:

At 7:26 PM, Blogger Chris Needham said...

I know you guys said it one of your posts, but which season's values are you using?

I'm assuming '04 NL? Or are you using combined MLB?

 
At 8:51 PM, Blogger DM said...

We are using the average of the past three years MLB (02,03,04), as provided on the Baseball Prospectus site.

 
At 1:19 AM, Anonymous John said...

Great FAQ on ERV scoring. Thanks again for providing this info. It would be interesting to see this line graphed across games to watch and compare different Nats players' game contributions against each other and over time.

Since you made a little dig in against keeping track of pitch count, I thought I'd make the argument that actually it's a very interesting stat to be recording and following, watching how the pitching battle and patterns evolve against different hitters.

Considering the week mid-reliever pitchers most teams seem to have these days, there's also the argument that drawing a lot of pitches--exercising some plate discipline and wearing out starters faster in the process--should itself arguably be a value contributed in an at bat.

 
At 11:37 AM, Blogger DM said...

John, you're right about the importance of pitch count -- the dig was just a personal view of mine after scoring it all these years and asking myself "Why do I do this?" But a lot of parks and broadcasts give you this info now, which is another reason I don't chart it anymore.

As for reflecting the value of working a lot of pitches, if that is true, then you should see that value come out in later at bats as the pitcher gets tired and gives up more bases. It would be hard to isolate that value, I think.

 
At 12:19 PM, Blogger Yuda said...

I occasionally chart pitches -- maybe once every third game or so -- at RFK since the scoreboard people can't be bothered to tell us (although it would undoubtedly be wrong even if they did), but I never do it anywhere else.

 
At 5:29 PM, Blogger Xeifrank said...

*However, the error means it is now 0.93 (Runner on first, no outs). Therefore, the error cost the team in the field 0.65 runs (0.28 minus 0.93). You can assign that value to the fielder, and give the batter the RV he deserved (-0.25, or 0.53 minus 0.28) because the play should have been made.*

Does this example lead to a zero sum game? In this example, nobody is given any positive value. The fielding team loses 0.65 and the hitting team loses 0.25 I understand why you did this, but it seems like if you want a "0 sum game), you need to have the RV of team 1 + the RV of team 2 equal 0 for each game state change.
Thoughts?
vr, Xeifrank

 
At 1:44 AM, Blogger DM said...

The FAQ left out the fact that the pitcher gets the positive RV of +0.25. Also, the remaining RV represents the runner on first -- if the pitcher gets out of the inning without letting him score, he will get that difference to his credit. You've heard of a pitcher getting "four outs"? ERV scoring accurately reflects that. A pitcher with a bad defense behind him but who still doesn't give up runs will have a higher RV.

I hope this clears it up.

 
At 9:47 AM, Blogger DM said...

Also, I do what the accountants do: make up a category to put stuff in. I use Team Running and Team Fielding for those values that are not properly assignable to any one person. In this case, Team Running gets the 0.39 value. The easiest example is a balk -- the runner does not deserve running credit, so it goes to Team Running.

 
At 6:49 PM, Blogger tüp bebek said...

thankssssss adminnnnnnnn niceeeeee siteeeeeeeeeeeeeeeeeeeeeeeeeee
tekirdağ koltuk yıkama
tekirdağ halı yıkama
tekirdağ halı yıkama
tekirdağ halı yıkama
tekirdağ çilingir
okey sitesi açmak

tekirdağ halıyıkama
ankara halı yıkama
tekirdağ temizlik
ankara tel örgü
mankenler
cast
su pompası
ısı yalıtım
kilo vermek

 
At 6:39 PM, Blogger dewa1111 said...

Thank you. Ankara Halı Yıkama

 
At 3:49 AM, Blogger Murat said...

Thanks nice blog.
Ankara Halı Yıkama, Çankaya Halı Yıkama, Çayyolu Halı Yıkama, Dikmen Halı Yıkama, Batıkent Halı Yıkama, Eryaman Halı Yıkama, Ankara Koltuk Yıkama, Çankaya Koltuk Yıkama, Çayyolu Koltuk Yıkama, Dikmen Koltuk Yıkama, Batıkent Koltuk Yıkama, Eryaman Koltuk Yıkama

 
At 2:28 PM, Blogger Murat said...

Thanks nice blog.
Ankara Halı Yıkama, Dikmen Halı Yıkama, Çankaya Halı Yıkama, Çayyolu Halı Yıkama, Batıkent Halı Yıkama, Eryaman Halı Yıkama, Ankara Koltuk Yıkama, Dikmen Koltuk Yıkama, Çankaya Koltuk Yıkama, Çayyolu Koltuk Yıkama, Batıkent Koltuk Yıkama, Eryaman Koltuk Yıkama, Ankara Halı Yıkama, Dikmen Halı Yıkama, Çankaya Halı Yıkama, Çayyolu Halı Yıkama, Batıkent Halı Yıkama, Eryaman Halı Yıkama, Ankara Koltuk Yıkama, Dikmen Koltuk Yıkama, Çankaya Koltuk Yıkama, Çayyolu Koltuk Yıkama, Batıkent Koltuk Yıkama, Eryaman Koltuk Yıkama, Ankara Halı Yıkama, Ankara Koltuk Yıkama, Halı Yıkama, Koltuk Yıkama, Çankaya Yıkama

 

Post a Comment

<< Home