In a previous life, I was pretty committed to being a scientist. Well, an engineer actually. A minor distinction to some, but the cardinal sin is to confuse the two among a certain group of my friends. I entered college bright-eyed and bushy-tailed, ready to learn what I needed to so I could go work for an Italian or Japanese MotoGP team making the next generation of fast motorcycles. Fast forward a few years to a bitter college sophomore curled up in the fetal position broken under the weight of abstract math and an electronics class that beat me up and took my lunch money . . .
Sitting in front of my advisor signing the paperwork that would transfer me to the business administration department, I heard something that stuck with me forever. “Tyler, you would have been a fantastic engineer. You’re just a very poor engineering student.” And with a flourish of the pen, that dream went poof. It turns out that the chalkboard wasn’t really for me. I really only did well in labs and practical applications of classroom work. Proof of this is my name on the wall for winning a truss-building competition and an academic record that reflects needing two attempts to get through Calculus 2.
Luckily, TTAG has provided me a really solid outlet to scratch the itch for the rigors of scientific testing. It further pleases me that my former engineering classmates spend a great deal of time in front of Microsoft Excel calculating budgets and not out in the field building, testing, and collecting data.
Recently, I posted an article detailing The Rehn Test, a standardized way of viewing pistol performance. As usual, there was a lively debate in the comments section. It gave me a lot of feedback on what was right and what was wrong with the way I was approaching this. But you can only sit in your living room arguing on the internet for so long before you have to take action. So I grabbed a bone stock M&P 9 that I recently acquired, an Apex Duty Carry Enhancement Kit, and a set of Trijicon sights. Figuring that the trigger and the sights are two known “weak” points of the M&P 9 pistols, I figured it might be fun to take a somewhat mundane review and add a bit more scientific rigor to it.
A few changes to the overall protocol had to happen first. Several of our commenters pointed out that if I wanted to be truly scientific, I should lose the holster draw portion. I agree completely, so all my testing was done with the gun indexed to the chest. Second, I declined using tests four through seven as I felt they were better suited for testing elements of the shooting process than sight acquisition and split times. For this test, I used tests one through three which I felt best evaluated sight acquisition and fast shooting.
My data collection methodology was fairly straightforward. Using a Competition Electronics shot timer, and starting from an indexed point on my chest, finger on the trigger, I attempted to shoot all A Zone hits. After the string was over, I recorded where my shots hit along with my time to first shot, and split times between shots. I started by running the stock M&P through ten repetitions of drills one, two, and three. Then I swapped out the trigger, ran it again, and then made the change to the sights running the same drills again. Once I got home, I put the data into Excel, and scored each rep using Comstock scoring.
The scoring is as follows. 5 points for A zone hits, 3 points for B/C zone hits, 1 point for D zone hits. You then sum all the points and divide that by the total time to get a normalized score. A higher Comstock score is better. During this process, there was a variety of interesting data that started to trickle out. Below is a screenshot of Excel with the relevant and interesting data points bolded for the first drill.
I really wish I’d been able to collect more data, but our budget is not endless and each one of those data points costs about a quarter. Though limited in data availability, some interesting data emerged. For the first drill with the completely stock gun, the ten runs yielded an average time to first shot of 1.06 seconds with the second shot coming at .32, the third at .30, and the fourth .30 seconds later. This yielded an average total time of 1.97 seconds for the drill. My average raw score was 19.40 (perfect is 20), and my average Comstock score for the stock gun was 10.01.
Once I installed the Apex Trigger, my average time to first shot went down below one second to .98, my second, third, and fourth shots coming .26, .31, and .23 later yielding an average time of 1.79. However, my accuracy suffered as my average raw score dropped a point to 18.40. This loss is accuracy was offset by the drop in speed as my Comstock score went up ever so slightly from 10.01 to 10.36, an increase of 3.5%
Making the final parts swap for the Trijicon Night Sights, my times dropped even further with my first shot coming at .96 followed by .27, .29, and .23 for the rest of them. This yielded an average total time of 1.75 AND my raw score went up to 19.14. Still not as accurate as the stock gun, but definitely an improvement. With the increase in accuracy and the very slight decrease in total time, my average score broke 11 with a score of 11.01. This represents a 10% increase in score over stock and a 6.3 % increase over stock sights with the Apex Trigger.
You’ll also notice that I bolded the standard deviation rows for each data set. As Nick has explained in his ammo consistency testing, standard deviation is one of a few good ways to assess the overall consistency of the data. A lower number is better, and while standard deviation isn’t the end all be all of statistical formulas, it does give a quick and dirty look at how “good” the data is. Broadly speaking, all of the standard deviation numbers are smaller with the Apex trigger and Trijicon sights in place over the stock trigger.
As a sidenote, you’ll notice that the Apex Trigger + Trijicon Sights group of data only has seven reps associated. This was due completely to a dwindling supply of ammo. I made the decision to lower the total reps for drill three so that I could collect as much data on all three drills as possible.
All of this data was very encouraging as it seemed to validate that spending the money to install an aftermarket trigger and set of sights would result in an increase in performance. The counter argument that familiarity with the course of fire is responsible for the increase in score is easily just as possible. The unfortunate fact is that eliminating that bias would take much more ammo than I had available for this test. I figure that running twenty five reps would be a good place to start with the goal of reaching a point where the accuracy and time numbers plateau. It wasn’t always rainbows and kittens as the data from the second drill proved.
With the stock gun, my average time to first shot was 1.14 and my second, third, and fourth shots came at .36, .41, and .36 yielding an average run time of 2.27 seconds with a standard deviation of .30. My average raw score with the stock gun was 17.40 (20 is perfect) with a standard deviation of 3.41. This yielded an average Comstock score of 7.70 with a standard deviation of 1.46.
Switching out the trigger yielded a slightly better raw score of 17.60 with an even better standard deviation number of 2.07. However, the average time to run the course rose to 2.62 seconds and yielded a Comstock score of 6.82, an 11.43% decrease in score. This increase in time was driven completely by the difference in split times between the two runs as the time to first shot remained almost identical.
The addition of a set of sights didn’t help either as the average total time remained identical (CONSISTENCY!), but the raw score number fell from 17.60 on average to 17.29. The standard deviation numbers shuffled slightly but overall the addition of Trijicon sights seemed make things slightly worse.
Where drill two was an unmitigated disaster in terms of shattering my preconceived notions, drill three seemed to be a blend of both. With the stock pistol, my time to the first shot average 1.21, my second, third, and fourth shots coming at .71, .67, and .74 seconds later to yield an average run time of 3.34 with a standard deviation of .24. My raw score was nearly perfect on average at 19 with a standard deviation of 1.94.
With the installation of the Apex Trigger, my run time got slower on average by .06 seconds with my first shot coming at 1.19, my second, third, and fourth coming at .76, .82, and .73 for an average time of 3.50 witha standard deviation of .30. My raw score also fell to 17.80 on average with a standard deviation of 1.99. This yielded a drop in score from 5.73 with the stock gun to 5.11 with the Apex equipped gun.
The addition of the Trijicon sights seemed to make things better as the average time to first shot, and split times fell bringing the average run time to 3.25 seconds with an average raw score of 18.86. This yielded a Comstock score of 5.83, a 1.7% increase over the stock gun.
Final Conclusions
If nothing else, usage of this test provided a more defined way to tell you that this gun with an Apex trigger and Trijicon sights seemed to be an absolute machine at “fighting distances” but was a bit harder to work at longer distances and yielded no significant difference at transitions back and forth from five to ten yards. It also made me as a reviewer take a step back to reconsider my previously held opinions. I left the range thinking that my performance at the ten yard line just needed a lot of improvement when the data showed me that the gear I was using was pretty consistently having a negative impact on my performance.
What this also showed me is that we can only really start to tease out story the data wants to tell us doing ten runs of a test. Realistically, something like twenty five runs (100 rounds) should be the norm to establish a plateau in performance. Either way, my hope is that this has added something to your knowledge bank as mine increased substantially over the course of a day out at the range with a shot timer and a copy of Excel.