12 replies to this topic

[CubChymyst]

So yesterday I commented that Grace was a better hitter than Sosa and made an attempt to defend it. Drew7 used sabermetric arguments to say I was wrong. Not being very familiar with sabermetrics I decide to look into them and I just want to make sure I've got this correct.

BA - does not tell the whole story because a hit depends on luck and the placement of the 9 players in the field (which makes me even happier about Dale defensive shifts). Which is why small sample sizes are not useful. Over the course of a season luck tends to average out and people with better contact skills tend to have higher averages.

OBP- is better because it speaks more of players overall ability to get on base which is one part of run scoring.

SLG- Talks about the ability to move players who are currently on base over (which is the second part of scoring runs). The basic thought is the more bases the batter advances the more bases current runners on base advance, and the greater the likely hood they will score. So a double is better than a single and a triple is better than a double.

I looked up how often a player scores from 2nd base on a single and I found 58% of the time he will score (not sure about that number the site I found didn't strike me as a primary source). It is safe to assume a runner will score from second 100% of the time on a double or better. So if I look at ability of 3 different hypothetical players to score a player from second base.

Player A, a 0.300 hitter who only hits singles In a 100 at bats he will have 30 singles drive in player from second (0.3 x 0.58) 17.4 % of the time.

Player B, a person batting 0.250 and slugging 0.350 and his numbers are 19 singles, 4 doubles, and 2 home runs in 100 at bats. He scores the player from second (0.19 x 0.58 + 0.04 + 0.02) 17.0% of the time. He knocks in the runner a lower percentage of the time but he also scores himself 2% of the time so he ends up with more RBI's. (plus he has the potential to knocks in anyone on 1st base)

Player C, a person batting 0.250 and slugging 0.490 and his numbers are 14 singles, 4 double, 1 triple, and 6 home runs in 100 at bats. He scores the player from second (0.14 x 0.58 + .04+.01+.06) 19.1% of the time and also scores himself 6% of time so he also has more RBI's.

(needed the math for myself)

So slugging makes up for the difference in average because extra base hits have a high chance of scoring the a run even if they occur less often. So do I have this correct?

[FFP]

The Force is strong in this one.

[Tommy]

Yeah, you can find a statistic to back up just about anything you want. No matter what you say, you'll always find someone that will have a differing view point. That doesn't make you wrong.

Throw sabermetrics out the window, man. Grace had an outstanding batter's eye, rarely struck out and was one of the best contact hitter's the Cubs have ever had. There are plenty of people (experts included) that would agree with your assessment.

[CubChymyst]

I completely agree that Grace had an awesome batting eye and better contact skills, and I don't think anyone disagreed with those two points. I just wanted to make sure I understood the opposite point better. The point with Sosa was that even though he had less hits and got on base less his hits were more likely got score an RBI. I'll agree that Sosa hits were more likely create RBI's, however I still think Grace was the better hitter for his career.

I went on ahead plugged the numbers for Sosa an Grace over their careers into the situation I used above (I know its not perfect but the general idea is 9th inning runner on 2nd) by taking the number of each type of hit and dividing by by total number of PA.

Grace - singled 21.3%, doubled 6.3%, tripled 0.5% and home runs 2.1% of his at bats (career BA 0.303, SLG 0.442)

Sosa - singled 15.6%, doubled 4.3%, tripled, 0.5% and home runs 6.9% of his at bats (career BA 0.273, SLG 0.468)

Grace scores the guy from second 21.3% of the time and Sosa 20.7% of the time, but Sosa ends up with more RBI because he scored himself 6.9% of the time vs Grace 2.1%. If it is the 9th inning and the game is tied Grace wins it more often, if the cubs are down by 1 Sosa wins it more often while Grace forces extra innings. As was mentioned before the difference doesn't end up being that large when you take their full career into account.

[Cubbie Blues]

Just wait till you learn about OPS+ or even better wRC+

[CubChymyst]

Just wait till you learn about OPS+ or even better wRC+

I'll put those two on the list for the next two things I learn. One thing I should of mentioned before is the winning situation I proposed only works in 1 run games and Ideally your going to avoid those by scoring more runs earlier.

[SirCub]

It's really interesting to hear your take on comparing the value of players by delving into how likely each of their hits is to contributing a run, and then looking at it in context of game situations and how those runs contribute to wins. This is absolutely the idea behind sabermetrics, but I like the way you go about thinking about it.

wOBA essentially takes the numbers that you did (how often a player scores from second on a 1B, 2B, 3B, HR) and applies it to every situation (nobody on, bases loaded, etc.) and looks at the probability that each one of those plays results in a run (or rather how many runs each is expected to result in based on their probabilities). That way you can weight each hit according to its run value, and they also incorporate the run-value of walks. Using the weighted value of each hit and walk, you can calculate a weighted On Base Average (combination of OBP, BA, and SLG, but better!).

[CubChymyst]

When I read about wRC+ during my lunch I also had to read about wOBA because that is the basis for wRC+ apparently. I really like that wOBA gives singles more weight than walks because with a hit there is a greater opportunity to drive in a run. Also with wOBA the fact that doubles are not weighted twice as much as a single was good to see because a runner can advance more than the number of base the hitter does.

What I'm picking up is sabermetric stats make an attempt to generalize a hitter over the course of a year by eliminate any the fluctuation in the position of runners when they come up to bat. Since different situation occurs somewhat randomly during a game (e.g. runner on second, bases load, or bases empty (I'm guessing coming up with the bases empty is the most common situation)) the idea is just to find the hitters that are most likely to generate RBI given all possibilities.

[SirCub]

Yea, pretty much you want to isolate the hitters contribution, without taking into consideration things like how good the hitters in front of him are at getting on base. One distinction is that it isn't really all about a hitters ability to knock in runs with RBI's but also his ability to get on base and potentially come around to score himself. Obviously, these are the two main components a hitter can make towards offense.

[CubChymyst]

That goes along with the one thing I read, don't remember where right now, is that the goal of a batter is to get on base and drive in the players in front of him. Since every inning starts with no one on base I'm assuming a batter generally comes up with no buddy on base and the more bases they can cover on their hit the better. For a first batter of an inning a walk and a single have the same result of placing a runner on first. Getting a double or triple makes it easier for the following batter to drive him in.

[Luke]

wOBA is one of my main go-to stats. It isn't perfect, but I would love to see it replace standard AVG for most purposes.

[fromthemitten]

here's some more advanced metrics: http://www.platoonad...son-voting.html

[CubChymyst]

here's some more advanced metrics: http://www.platoonad...son-voting.html

Favorite part about that was seeing Darwin Barney as a likely MVP candidate using Archaic statistics.