Quantifying the Impact of K and BB Rates on Offensive Production

When evaluating a hitter or pitcher’s statistical performance, the vast majority of what you need to know comes from the following rate data – strikeouts, walks, popups, line drives, hard and weak flyballs, and hard and weak groundballs. Put some of those on the good side of the ledger, the rest on the bad, and pretty quickly one has a feel for the present ability, and along with scouting data, the future projection of the player. Today, let’s focus on strikeouts and walks, and on quantifying their overall impact on hitters’ offensive production. How much must a hitter do to compensate for poor K and BB rates, and much additional margin for error is provided by strong K and BB rates.

Our control group will be 34 position players who have changed clubs this offseason and logged substantial playing time in 2013. First, let’s take a look at these players’ 2013 strikeout/walk data expressed in a few different ways.


Last First K BB K % BB % K %:100 BB%:100 K PCTILE BB PCTILE GROUP
Aoki Norichika 40 55 6.1% 8.3% 30 103 1 55 1
Arencibia JP 148 18 30.0% 3.7% 150 46 95 4 4
Barnes Brandon 127 21 29.1% 4.8% 143 60 89 14 4
Beltran Carlos 90 38 15.3% 6.5% 75 80 27 27 3
Byrd Marlon 144 31 25.5% 5.5% 125 68 87 15 4
Cano Robinson 85 65 12.7% 9.7% 63 121 18 70 1
Choo Shin-Soo 133 112 19.4% 16.4% 94 199 57 98 1
Davis Rajai 67 21 18.9% 5.9% 95 75 53 20 3
Doumit Ryan 99 48 18.6% 9.0% 92 113 52 64 1
Ellis Mark 74 26 15.8% 5.5% 77 69 30 16 3
Ellsbury Jacoby 92 47 14.8% 7.6% 73 94 29 37 3
Fielder Prince 117 75 16.7% 10.7% 83 133 41 81 1
Fowler Dexter 105 65 21.7% 13.4% 107 167 71 92 2
Freese David 106 47 20.9% 9.3% 102 114 66 66 2
Ibanez Raul 128 42 26.1% 8.6% 129 107 81 56 2
Infante Omar 44 20 9.3% 4.2% 46 53 2 11 3
Johnson Kelly 99 35 24.6% 8.7% 122 109 76 58 2
Jones Garrett 101 31 23.4% 7.2% 115 89 78 36 4
Kinsler Ian 59 51 9.8% 8.5% 48 105 5 55 1
Lough David 52 10 15.8% 3.0% 78 38 35 2 3
McCann Brian 66 39 16.7% 9.8% 83 123 38 73 1
McLouth Nate 86 53 14.7% 9.1% 73 113 30 65 1
Morneau Justin 110 50 17.6% 8.0% 87 100 44 49 1
Morrison Logan 56 38 17.1% 11.6% 86 146 45 89 1
Murphy David 59 37 12.5% 7.8% 62 98 16 46 3
Peralta Jhonny 98 35 22.0% 7.8% 110 99 68 47 4
Pierzynski AJ 76 11 14.8% 2.1% 72 26 27 1 3
Ruggiano Justin 114 41 24.8% 8.9% 121 110 85 64 2
Saltalamacchia Jarrod 139 43 29.8% 9.2% 149 116 94 67 2
Schumaker Skip 54 28 15.4% 8.0% 78 102 31 53 1
Smith Seth 94 39 23.4% 9.7% 115 120 71 68 2
Stubbs Drew 141 44 29.7% 9.3% 147 116 90 66 2
Trumbo Mark 184 54 27.1% 8.0% 136 101 85 49 2
Young Chris 93 36 25.1% 9.7% 126 123 79 70 2

The table above lists each player’s raw K and BB totals, their K and BB rates, those rates scaled to MLB average of 100, and expressed in overall percentile rank within the MLB regular player population. The players are also placed in groups (far right column) as follows: Group 1 = K rate below MLB average, BB rate above MLB average; Group 2 = K and BB rates below MLB average; Group 3 = K and BB rates above MLB average, and Group 4 = K rate above MLB average, BB rate below MLB average. This table alone, however, tells us a limited amount about each player’s offensive ability. Let’s fill in the picture a little more by adding performance on balls put in play (including homers) for each player, and then quantify the impact of the K and BB data on each hitter, and each group of hitter.

Last First BIP AVG BIP SLG BIP RUN BIP R:100 TOT AVG TOT OBP TOT SLG TOT RUN TOT R:100 GROUP
Aoki Norichika 0.299 0.387 3.54 72 0.279 0.339 0.361 3.88 100 1
Arencibia JP 0.284 0.532 4.54 93 0.196 0.225 0.366 2.48 64 4
Barnes Brandon 0.336 0.484 4.91 100 0.233 0.270 0.337 2.79 72 4
Beltran Carlos 0.350 0.580 6.09 124 0.293 0.338 0.485 5.03 130 2
Byrd Marlon 0.397 0.695 8.28 169 0.290 0.329 0.507 5.02 130 4
Cano Robinson 0.358 0.592 6.36 130 0.308 0.375 0.508 5.79 150 1
Choo Shin-Soo 0.369 0.599 6.63 135 0.283 0.401 0.460 5.74 148 1
Davis Rajai 0.320 0.462 4.46 91 0.255 0.299 0.369 3.40 88 2
Doumit Ryan 0.303 0.487 4.43 90 0.241 0.310 0.388 3.68 95 1
Ellis Mark 0.317 0.412 3.99 81 0.264 0.305 0.343 3.27 84 2
Ellsbury Jacoby 0.350 0.499 5.28 108 0.294 0.347 0.419 4.50 116 2
Fielder Prince 0.341 0.563 5.76 118 0.278 0.355 0.458 4.96 128 1
Fowler Dexter 0.347 0.538 5.61 114 0.260 0.360 0.403 4.54 117 3
Freese David 0.334 0.490 4.94 101 0.257 0.326 0.377 3.83 99 3
Ibanez Raul 0.331 0.675 6.76 138 0.237 0.302 0.482 4.37 113 3
Infante Omar 0.348 0.494 5.20 106 0.314 0.343 0.446 4.68 121 2
Johnson Kelly 0.320 0.558 5.36 109 0.234 0.300 0.408 3.72 96 3
Jones Garrett 0.310 0.560 5.21 106 0.232 0.287 0.419 3.63 94 4
Kinsler Ian 0.301 0.443 4.02 82 0.269 0.331 0.396 4.05 105 1
Lough David 0.332 0.481 4.82 98 0.278 0.300 0.403 3.68 95 2
McCann Brian 0.313 0.564 5.30 108 0.255 0.328 0.459 4.57 118 1
McLouth Nate 0.307 0.475 4.38 89 0.258 0.325 0.398 3.99 103 1
Morneau Justin 0.315 0.494 4.68 95 0.254 0.314 0.399 3.84 99 1
Morrison Logan 0.299 0.466 4.19 85 0.241 0.329 0.376 3.86 100 1
Murphy David 0.252 0.427 3.23 66 0.218 0.279 0.369 3.14 81 2
Peralta Jhonny 0.399 0.601 7.22 147 0.304 0.359 0.457 5.03 130 4
Pierzynski AJ 0.320 0.500 4.81 98 0.272 0.287 0.425 3.68 95 2
Ruggiano Justin 0.308 0.551 5.09 104 0.224 0.293 0.401 3.57 92 3
Saltalamacchia Jarrod 0.407 0.695 8.49 173 0.274 0.340 0.467 4.82 125 3
Schumaker Skip 0.313 0.396 3.80 78 0.261 0.320 0.329 3.36 87 1
Smith Seth 0.338 0.524 5.32 109 0.251 0.323 0.388 3.88 100 3
Stubbs Drew 0.339 0.519 5.29 108 0.228 0.300 0.349 3.25 84 3
Trumbo Mark 0.327 0.636 6.27 128 0.231 0.292 0.449 3.94 102 3
Young Chris 0.277 0.525 4.38 89 0.200 0.278 0.379 3.20 83 3
MLB AVG 0.323 0.505 4.90 0.253 0.318 0.396 3.87

An editorial comment – it’s always bugged me that HR are excluded from BABIP calculations. Not all HR are created equal – a ball that goes over the fence and is excluded from BABIP for one hitter, goes into a glove and is included in BABIP – negatively – for another. End of editorial. Also, for the purposes of this exercise, the BIP and Total (including K and BB) run values are being calculated and scaled to the MLB average ERA as follows: the square of ((1.7 * Player OBP + Player SLG)/(1.7 * MLB OBP + MLB SLG)) * MLB Avg ERA. Also, I am arbitrarily excluding HBP from the above OBP calculations, and including SH and SF as ordinary outs for the purpose of this exercise. The estimated run values excluding and including the K/BB data are scaled to 100 in the 4th and 9th columns above. They are not adjusted for park factors.

Based on performance on batted balls alone, Jarrod Saltalamacchia (with plenty of help from the Green Monster) and Marlon Byrd (with some help from the surprisingly compliant Citi Field LF/LCF area) are the most productive, with the somewhat unlucky David Murphy and groundball machine Norichika Aoki bringing up the rear. The average MLB player’s run value scaled to ERA is decreased by 1.03 (from 4.90 to 3.87) by the addition of the K/BB data – the 34 players above saw their run values decrease by an average of 1.22. Now let’s look at the results by K/BB group.

GROUP # AVG -RUN +/- R100
- K, + BB 1 0.49 13.6
- K, – BB 2 0.81 4.8
+ K, + BB 3 1.84 -16.2
+ K, – BB 4 2.24 -25.2

The table above shows how significantly a high K rate, especially, negatively impacts a player’s ability to achieve a high level of production. Hitters in Group 3 – higher than average K rate, higher than average BB rate – see their run value scaled to 100 fall by an average of 16.2 basis points from their run value based on batted balls alone. The Group 4 hitters – higher than average K rate, lower than average BB rate – are hit even harder, by an average of 25.2 basis points. This is the dynamic that drives a Mark Trumbo – a Group 4 player with a batted ball run value of 128, into a league average-ish 102 guy once K and BB are added in. It makes someone with an acceptable batted ball profile like JP Arencibia (Group 4 – 93) into a sub-replacement level offensive player (64). It makes someone like Marlon Byrd (Group 4 – off-the-charts 169 on batted balls) utterly reliant on such a level of batted ball authority because of the huge minus contributed by his poor K and BB rates. Conversely, players with relatively unimposing batted ball profiles like Norichika Aoki (Group 1 – 72) and Ian Kinsler (82) whose Group 1 K/BB profiles make them league average-ish offensive players with relative run values of 100 and 105 once K and BB are taken into account. In a nutshell, if you’re going to strike out a lot, you had better punish the baseball consistently when you do make contact.

The tables above give us more reasons why a hitter like Robinson Cano is as good as he is. Based on his batted ball performance alone, five hitters above rank above his relative run value mark of 130. His solid Group 1 K/BB performance bumps him way up to 150 in terms of overall relative run value, the best among this group. Cano isn’t overly reliant on impact batted ball authority, or on impact K or BB rates. He’s simply good at all of these aspects of hitting, and while each of his abilities will deteriorate over time, he’s unlikely to suffer a sudden plummet because of a breakdown in any single aspect of his game – his other skills will keep afloat longer than a player overly reliant on one single component.

It is self-evident that K and BB rates impact the production of position players and pitchers, but the above helps quantify that impact. A very similar analysis for pitchers can and will be prepared in an upcoming post. Obviously, there are many angles not taken into account here – park factors, the impact of player speed on performance on balls in play, luck, etc. – but this much is clear. Maintaining solid K and BB rates do not alone make a hitter good or bad, but doing so clearly enhances the margin for error a hitter possesses with regard to the authority with which he puts the ball in play. Ian Kinsler’s batted ball authority has declined in the recent past – but his K and BB rates have allowed him to at least continue resembling Ian Kinsler – for now.




Print This Post





9 Responses to “Quantifying the Impact of K and BB Rates on Offensive Production”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Cody says:

    “An editorial comment – it’s always bugged me that HR are excluded from BABIP calculations. Not all HR are created equal – a ball that goes over the fence and is excluded from BABIP for one hitter, goes into a glove and is included in BABIP – negatively – for another.”

    I’ve never thought of that before and I love the observation.

    Vote -1 Vote +1

  2. tz says:

    Real cool stuff.

    On your editorial on BABIP, I think the primary driver was the development of fielding-independent numbers, where excluding HRs makes sense.

    For this type of analysis, however, it makes sense to include HRs. What we really have are three skills:

    – Drawing walks
    – Avoiding strikeouts
    – Quality of balls hit into play (including HRs)

    Again, this is great stuff.

    Vote -1 Vote +1

  3. Peter Jensen says:

    It is self-evident that K and BB rates impact the production of position players and pitchers, but the above helps quantify that impact.

    Does it really? I found this article to be one of the most confusing I have ever read. Does it really offer more than “more walks are good and fewer strikeouts are also good”? If you want to correctly value offensive production stick to linear weights.

    Vote -1 Vote +1

  4. channelclemente says:

    I LOVE this article. Thanks.

    Vote -1 Vote +1

  5. dcs says:

    I kind of just scanned the article. But it seems to me that he is making the mistake of holding the batted ball results as independent of the K rate. In reality they are intimately intertwined, as harder power-type swings result in both Ks and harder hit balls.

    Vote -1 Vote +1

    • Chris from Bothell says:

      I was kinda wondering where your average three-true-outcomes sort of guys sit, in all this.

      Vote -1 Vote +1

    • tz says:

      Good point. I think that there are two items that interact here:

      – For any individual player, there is probably some correlation between K rate and harder hit balls reflecting a change in approach. (see Byrd, Marlon 2013)

      – Across players, there is probably a more independent relationship between K rate and power, simply because of the different skill sets (i.e. not everyone can be Miguel Cabrera)

      Vote -1 Vote +1

  6. DrBGiantsfan says:

    It looks to me like there are good and bad hitters in approximately equal ratios in all 4 groups. Maybe K’s and BB’s don’t matter as much as we give them credit for?

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>