The 20-80 Scouting Scale and what the numbers all mean

Often times we hear things thrown around like “he has a 50 arm” or he has “80 speed”. But what do those things mean exactly? Hopefully, I can explain what they are meant to tell us and provide some examples as well.

The Scale

The scouting scale is from 20-80. I still haven’t quite figured out why, but have heard others speculate that when it first started that they decided 50 would be average (which assumes they likely wanted to use 0-100). It is speculated that someone then said they should use the scientific scale of three standard deviations above and below average. So if 50 were average, then three grades in each direction would be the normal distribution.

So, for those of you non-math guys, to make it easier, let’s look at it in a much easier way. Essentially, the talent is distributed like that of a bellcurve, with the peak of the curve being average. In a normal bell-curve, here is how it breaks down from a percentage standpoint:

Now, when we look at the chart above, 68.2% is an awful lot of players. That is why more teams use the half-grade as well. That makes 45-50 worth about 34.1% and 50-55 another 34.1%. Basically though, most players fall into the average range. When it comes to Major Leaguers only, there is roughly only 1 or 2 players that would fall into the 20 or 80 range if we go with the pure mathematical breakdown listed above, though if we truly wanted to get into the numbers there could be a few more depending on just how varied the numbers were.

Real Life Application

While scouts have a general rule of thumb of what each tools breaks down to (IE: 90 MPH is average for a fastball, 40 HR is plus-plus power), those things aren’t always based in reality as the talent level does change over time. So, what I am going to do is break down the different tools based on available statistics and put the above rates on them to show what each tool would grade out as and provide an example of it.

The tools for position players are as follows: Hitting, power, speed, defense and throwing arm. I will look at these first.

Hit Tool – Measures the ability to hit for average.

Power – Measures the power, typically home run power of a player.

Speed – Measures the speed of a player.

Defense – It measures the defensive ability, reflective of position. A guy who is say a 50 level shortstop would probably receive a better grade at second base.

Arm – Measures the arm strength, not reflective of position.

Now that we explained what each one is supposed to reflect, let’s put it to use.

What I was planning on doing was taken care of yesterday afternoon by Mark Smith at Fangraphs, who apparently thinks just like I do. He looked at batting average and came up with these numbers:

That looks right to me. Miguel Cabrera had the highest average in the game during the 2010-2012 stretch by a large stretch over Joey Votto who came in with a .321 average and second place on the list. Votto would be a “75” hit tool player according to the true breakdown.

Getting away from what Mark had, I want to look at power a little differently than he did. He looked at IsoP, which is a good start, but I personally would rather look at home runs. A guys speed can lead to a higher IsoP than his true power would lead you to believe versus a slow guy. While I would take it a step further, and do something like HR/(AB-K) then adjust for park factors, pure home runs per 600 at bats sounds good (some guys get more chances to swing away because of where they bat in the lineup, so I am normalizing).

For the most part that looks right. Those numbers vary a little bit from what scouts actually use (for the most part). They use something that is similar to this: 80 (39+), 70 (32-38), 60 (25-32), 50 (17-25),  40 (11-17), 30 (5-11) and 20(0-5). It is similar  until we get to the 60 and up range.

Speed can’t really be shown with a stat. Steals don’t represent speed. Some guys are just good or bad base runners. Speed is usually rated by a hitters time to first base.  They are timed from the point of contact until they reach first base (non-rounding first base event). Here is the scale (bunts do not count):

I don’t have examples for each range here.

Defense is another thing that you can’t put a real number on, at least not yet. I think that one day it will be possible with the new system MLBAM is using with the Field F/X system. But things such as range and glove go into this, as well as pitch framing and agility with catchers.

Finally there is the arm. It is another thing that you can’t put a real number on. Using assists doesn’t do us much because guys with known strong arms won’t be run on as much. Not all opportunities are created equally either. A guy like Ichiro, in his prime, would have been an example of an 80 arm. David Eckstein was probably a 20 arm.

Now with pitchers, things are more simple. Each pitch they throw is graded on the 20-80 scale, as well as control and command. Control is the ability to throw strikes. Command is the ability to locate within the strikezone.

Every pitcher has a fastball, even a knuckleballer like R.A. Dickey has a fastball that he will use from time to time. Fastball velocity is generally used when ranking a fastball, though there will also be notes on the movement and deception of the pitch. However in terms of pure velocity, here is how the actual stats from 2010-2012 played out (note that I broke it down for guys with less than 50% of their appearances as starters and more than 50%) and be sure to note that they are the average velocities and that plenty of guys can reach several MPH higher than what they average:

Now obviously, Aroldis Chapman and several others can throw 100+, even with relative frequency, but he led all of baseball in that time with an average velocity of 98.1 MPH. The chart above is for the full three standard deviations above average, so no one qualified as an “80”. But obviously, guys can get there on any given night. Also, be sure to note that left handers generally get a MPH taken off of what the average is for a right handed pitcher to achieve the same grade.

With other pitches plenty of things go into what grades the pitch. Breaking balls are velocity, bite, depth, arm action to sell the pitch. Change ups are arm action, separation from the fastball (in velocity) and movement. Of course, being able to throw that pitch where you want it can make it play up from the raw grade that a pitch could be given too.

Control is not incredibly tough to put a number on now that we have Pitch F/X data that we can use. For this, I simply used “zone percentage” which is the amount of time that a pitcher throws a pitch inside of the strikezone.

I will be honest and say that Mariano Rivera being listed there is downright shocking given that his walk rate is among the league best of 1.5 per 9 innings in that time. But, the guy simply doesn’t throw that many pitches in the zone apparently. Where as another guy with an incredibly low walk rate, Cliff Lee (1.2 walks per 9 innings) pounds the strikezone better than any pitcher in the league.

Command is tough to put a number on because we would need to chart every game a pitcher has because we need to know the location of the pitch and the location where the catcher called for the pitch to be at. We don’t have access to anything like that. Plus, like control, a pitcher can have good control/command of one pitch and not so much of another. So this could vary for each pitch, though an overall grade will still be given and is likely tied to the fastball control/command since most pitchers throw that pitch at least 60% of the time.

Statistical “Scouting Scale” by pitch

For this, I thought it would be good to look at the Pitch Values from Fangraphs (which are based on the actual results of each pitch) to share who got the results from each pitch that fall into the different scouting ranges. I am using the value per 100 pitches thrown (of that specific pitch). A player also had to throw at least 5% of that pitch to be eligible. If there was a Red in the range of the grade, I tried to place them as the example.

Hopefully after all of that, I didn’t confuse anyone more than they were coming into the article. If you have any questions, or comments leave it in the comments and we can all talk about it.