SLG and Speed

by 5Tool on Jul 2, 2009 11:11 PM EDT

Who would you rather have in your lineup, a fast player who can extend many a single to a double or a slow player, who regardless, has a the same SLG as the fast player?

The first player cannot advance runners as well, but is more easily advanced. The second player advances runners well, but cannot be advanced as easily.

Obviously one would be a better clean up hitter and the other a lead off hitter, but for the moment my question is, who do you think is worth more runs per game assuming you have 8 other average batters on the team? Offense only please.

Tweet Comment 13 comments | 0 recs |

Do you like this story?

Comments

Interesting question.

Something I’ve been thinking about lately myself. The way you worded it makes it a no-brainer though – if they have the same SLG, why not take the speed?

OverTheMonster - ALLERGEN WARNING: May contain peanut butter.

by bdalebs on Jul 2, 2009 11:41 PM EDT reply actions

Exactly what I was going to say.

No mention of OBP, so I’ll take slug and speed.

If you were thinking, you wouldn't have thought that.

http://sonsofsamhorn.net/index.php?showforum=129

by Warden11 on Jul 3, 2009 12:01 AM EDT up reply actions

You might not take the speed because that dude's SLG doesn't actually advance other runners as much.

For example, some of Juan Pierre’s doubles are like the average person’s singles in advancing runners and some of Frank Thomas’ singles are like the average person’s doubles. In other words’ Frank Thomas’ “advancement SLG” (how much his SLG advances others’) is much better than Juan Pierre’s.

For a question like this RE24 might be a good tool to use, because it tells you how much others advance.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Jul 3, 2009 8:25 AM EDT up reply actions

Exactly

You completely understood my question Sky, but what exactly is RE24?

by 5Tool on Jul 3, 2009 3:55 PM EDT up reply actions

Run expectancy based on the 24 possible base/out states

For each base/out state (e.g. runner on 1st, 2 outs), how many runs would you expect to score (0.251 in this case).

You can measure the run expectancy before and after an at-bat to determine the value of that at-bat (and then give the credit to the batter).

Tangotiger has calculated the RE for all the base out states for 1999-2002 which is a good place to start. You’ll really want custom RE charts based on the run environment, but these will work for most things.

Win Probability Added (WPA) works almost exactly the same way, except using the chances of a team winning the game before and after the at-bat.

by Dan Turkenkopf on Jul 3, 2009 7:56 PM EDT up reply actions

And so it's useful to 5Tool's question because....

It will tell you about how far other runners advanced on a player’s hits. Now, it involves some “clutchiness” and other situational information you probably don’t want either. Which means if I think about it again, you’d probably want to run an original study to test how far runners advance on various hitters’ hits (adjusting for runner, park, and defense, natch.)

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Jul 3, 2009 9:07 PM EDT up reply actions

Oh...

Because the ball isn’t hit to a position optimal for runner advancement… Got it.

OverTheMonster - ALLERGEN WARNING: May contain peanut butter.

by bdalebs on Jul 6, 2009 3:52 PM EDT up reply actions

I'm curious of how to mathematically figure out Win Probablity

I tried googling it for a while and I didn’t get anything. Then Studeman implemented his WPA Inquirer using math. How did he do it?

Derosa.

by vivaelpujols on Jul 4, 2009 9:18 AM EDT up reply actions

Markov Chains I'm pretty sure.

by Dan Turkenkopf on Jul 4, 2009 1:54 PM EDT up reply actions

Maybe data from Tango?

With a linear assumption about how the value of events change given different run environments? Is there any explanation on the front page of that spreadsheet?

Actually, whether or not he got the data from Tango, you can build up the values just from having base-out run frequency data and win-probability data at each half-inning point. For example, if there’s one out and a runner on second, you can know the probability of 0,1,2,3, etc. runs scoring to the end of the inning. For each possibility, you look up the win-probability in a table. Do a sum of the probability of each outcome times its resulting win-probability.

On a side note, one definition of leverage is the standard deviation of the changes in win probabilities given the probability of each even (single, out, walk, etc.) at each game state. I think Tango’s favorite uses the weighted average of the absolute changes in WPA given each possible event.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Jul 4, 2009 4:59 PM EDT up reply actions