- FanGraphs Baseball - http://www.fangraphs.com/blogs -

Joey Votto and the Value of a Walk

Joey Votto is a fantastic player. There is absolutely no denying this fact. Any and all arguments regarding the quality of Votto’s offensive ability should be, and generally are, made in terms of whether he is simply an excellent hitter, or one of the very best hitters currently playing baseball today.

The argument “against” Joey Votto commonly refers to his “inability” to “drive in runners”. In other words, Votto had a low RBI total in 2013, given his salary, place in the lineup, and position. Votto’s 73 RBI ranked 17th in baseball among first-basemen, and 65th overall.

We all know that RBI is not an effective way of measuring offense, or even the ability to drive in runners, because of its dependence on the offensive quality of the rest of the team. To rectify at least this flaw of the RBI, let’s look at a variant of the statistic, one that is reflective of its purpose, but more effectively measures said purpose. We’ll call it RBI per  Opportunity, or RBI/Opp. Simply, it is the percentage of runners, on base while a player is up, that are driven in by said player, removing plate appearances in which the player is intentionally walked:

Num Player RBI Opps RBI/Opp
1 Wilson Ramos 48 196 24.5%
2 Allen Craig 87 359 24.2%
3 Freddie Freeman 91 382 23.8%
4 Chris Davis 85 376 22.6%
5 Miguel Cabrera 95 423 22.5%
6 Matt Carpenter 68 307 22.2%
7 Paul Goldschmidt 94 428 22.0%
8 Robinson Cano 81 374 21.7%
9 Salvador Perez 67 312 21.5%
10 Yadier Molina 71 342 20.8%
192 Joey Votto 53 412 12.9%

Note: these are not literal RBI, but the number of times the runners on base actually score. In other words, I’m counting runs that score as a result of double plays or errors as RBI.

The leaders in RBI/Opp are, for the most part, the players that we would expect to be leaders. These are primarily players that are, and should be, leaders in MVP voting. Joey Votto is a not a leader in this statistic. Joey Votto was not a victim of opportunity, a victim of the players who batted in front of him. Joey Votto drove in well fewer runners that we would expect from a player of his caliber.

Of course, that does not mean that Joey Votto should not be considered in MVP voting. There is more to being an effective offensive player than the ability to drive in runners. You know this. I know this. Even someone like Harold Reynolds knows this. However, the argument made by people like Reynolds is that whatever value Votto brings apart from driving in runners is minimal compared to the value he loses by failing to do so.

A simple way of posing this argument is that Votto is responsible for driving in those runners, that it is the job of Joey Votto, being a middle-of-the-order player, to drive in the runners that are on base when he is up, because the players behind him may not. Specifically, the implication is that a walk is of little importance with runners on base, because unless the bases are loaded, a walk does not immediately score runs. A walk just passes responsibility to someone else to drive in runs.

This is almost true, in a way. If we’re considering the context of plate appearances, then walks with runners in scoring position and first base open are surely less valuable, relatively, than walks in other situations. Walks with two outs are surely less valuable than walks with no outs. Certain events have more value in certain situations, and walks have less value in situations with more RBI opportunities, in general.

How far less? Consider the run value of each type of outcome based on the base-out state:

Outs Runners 1B 2B 3B HR BB K Out
0 ___ 0.37 0.60 0.91 1.00 0.37 -0.22 -0.22
1 ___ 0.24 0.40 0.67 1.00 0.25 -0.15 -0.15
2 ___ 0.12 0.21 0.25 1.00 0.12 -0.09 -0.09
0 1__ 0.68 1.12 1.53 1.63 0.59 -0.34 -0.38
1 1__ 0.44 0.95 1.42 1.75 0.38 -0.28 -0.31
2 1__ 0.22 0.70 1.14 1.88 0.20 -0.21 -0.21
0 _2_ 0.72 0.98 1.31 1.39 0.36 -0.42 -0.31
1 _2_ 0.64 0.98 1.26 1.60 0.22 -0.34 -0.33
2 _2_ 0.71 1.00 1.05 1.79 0.11 -0.31 -0.31
0 __3 0.40 0.64 0.95 1.03 0.37 -0.52 -0.40
1 __3 0.62 0.77 1.03 1.38 0.27 -0.52 -0.24
2 __3 0.80 0.89 0.97 1.68 0.06 -0.41 -0.41
0 12_ 1.01 1.59 2.00 2.09 0.91 -0.51 -0.47
1 12_ 0.85 1.53 2.01 2.34 0.61 -0.49 -0.56
2 12_ 0.90 1.63 2.01 2.75 0.39 -0.34 -0.34
0 1_3 0.68 1.14 1.58 1.66 0.48 -0.66 -0.54
1 1_3 0.78 1.30 1.78 2.10 0.38 -0.67 -0.51
2 1_3 0.96 1.45 1.87 2.62 0.26 -0.47 -0.47
0 _23 0.85 1.15 1.46 1.54 0.36 -0.57 -0.38
1 _23 0.92 1.28 1.56 1.89 0.17 -0.78 -0.47
2 _23 1.44 1.74 1.79 2.53 0.17 -0.57 -0.57
0 123 1.07 1.60 2.10 2.18 1.00 -0.76 -0.68
1 123 1.18 1.92 2.40 2.72 1.00 -0.78 -0.73
2 123 1.50 2.19 2.62 3.36 1.00 -0.74 -0.74

There’s a lot of information in the above table, and you don’t need to sift through all of it. The important thing to notice is that the values of each event relative to other states, as well as relative to other events within the state, vary greatly. A home run with no one on and two outs is worth more than eight times more than a walk in that situation. But a home run with a runner on first and no outs is worth less than three times as much as a walk.

To really understand the value of a walk relative to other events, we need to do more than just manually compare the values —  we need to standardize each state so that the average event in each is worth the same. If the average event in each base-out state is the same, then we can see more clearly when the walk is more and less important. Using a similar process as the one for calculating wOBA, we do this by subtracting the run value of an out from the run value of each event, then normalizing so that the average positive event in each base-out state is worth ‘1’, to mimic OBP.

When all is said and done, the relative “value” of a walk — or, the wOBA weight given the situation — in each base-out state looks like this:

Runners/Outs 0 1 2
___ 0.843 0.806 0.715
1__ 0.795 0.719 0.621
_2_ 0.667 0.582 0.446
__3 0.882 0.631 0.430
12_ 0.844 0.749 0.568
1_3 0.765 0.643 0.507
_23 0.647 0.514 0.452
123 0.879 0.802 0.713

You can see that in the states with runners in scoring position and first base is open, a walk, in general, has far less value relative to other events, especially as the number of outs increases. With two outs, runners on, and first base open, the value of a walk drops dramatically, whereas its value is almost average (1) relative to other events with no outs and bases empty or a man on third.

With these values in mind, let’s now look at where Votto’s 116 non-intentional walks came this season, along with the above relative value or wOBA weight of the walk in that state:

Outs Runners PA niBB niBB% BB wOBA
0 ___ 147 12 8.2% 0.843
1 ___ 92 17 18.5% 0.806
2 ___ 155 27 17.4% 0.715
0 1__ 30 7 23.3% 0.795
1 1__ 66 14 21.2% 0.719
2 1__ 43 7 16.3% 0.621
0 _2_ 17 5 29.4% 0.667
1 _2_ 39 5 12.8% 0.582
2 _2_ 9 2 22.2% 0.446
0 __3 2 1 50.0% 0.882
1 __3 13 4 30.8% 0.631
2 __3 7 0 0.0% 0.430
0 12_ 15 3 20.0% 0.844
1 12_ 14 3 21.4% 0.749
2 12_ 16 3 18.8% 0.568
0 1_3 7 0 0.0% 0.765
1 1_3 7 2 28.6% 0.643
2 1_3 6 0 0.0% 0.507
0 _23 2 0 0.0% 0.647
1 _23 2 0 0.0% 0.514
2 _23 4 0 0.0% 0.452
0 123 3 0 0.0% 0.879
1 123 6 0 0.0% 0.802
2 123 5 4 80.0% 0.713

It’s difficult to tell, based on this table alone, whether Votto walked in the “correct” situations or not. You can easily point out situations in which he walked too much given the value of a walk (runner on second, 1 out) and situations in which he did well in not walking when a walk wasn’t as important (runners on second and third).

We can determine whether Votto walked at the right times overall by taking the weighted sum of the wOBA weights of the walks in these situations. In doing so, we end up with .727. Using league-wide numbers and the above weights, the average wOBA weight on walks is .719. In other words, Votto walked in equally or more important situations than the average hitter. In other other words, Votto’s walks were not useless. Votto’s walks were about as useful as we expect walks to be.

This analysis can be extended to more than just walks. We can calculate these base-out wOBA weights for every event, applying them to all of Votto’s plate appearances in order to measure how well he hit to the situation overall.

When we do so, Votto ends up with a .403 base-out wOBA. His actual wOBA was .400. Just as with with walk, we find that Joey Votto’s hitting when considering the relative value of events in the base-out states was equivalent to Votto’s hitting using context-independent weights. Votto did not walk when he should have singled, or homered when a single would do. And lest you think Votto simply got fewer hits, instead of the wrong hits, in important situations, his RE24 on the season was 51 runs, compared to 49 weighted runs above average (wRAA).

Whether or not Votto’s lack of RBI in 2013 is predictive of a lack of RBI in future seasons, fans are right to be concerned that he did not live up to expectations this season. With no other information, one would be justified in believing that Votto did not do enough with runners on base, that he did not do enough to produce runs, that he walked too much when walks were far less valuable than hits.

One would be justified in believing this, but one would also be incorrect in believing this. Votto did not walk at the wrong times, and he did not hit more poorly in important run-scoring opportunities than otherwise. Votto’s run production taking context into account was as we would expect it to be without context taken into account: not just good, but one of the greatest in the game.