I know some of you guys have been following our comparison study. Both rifles now have 1200 rounds through them. We shot the second round of accuracy tests yesterday and have compiled all of the accuracy test data in the form of a table. Here is a link to that table-

Wall Photos | Facebook

Both rifles have fired 1200 rounds with no reported issues, except one 20 round Colt magazine was damaged during a reloading drill. The mags fall to the ground from about chest height. Whie reloading the magazine with cartridges, the shooter had difficulty getting the cartridges to stay in the magazine. The feed lips have obviously been damaged. We have pulled that mag out of the rotation.

thanks Matthew thats great data

Matthew ....... are the 883 and 095 some sort assigned of I.D. number for each rifle, and it the last header entry for the Colt a typo (the Sport's 095 instead of 883)?

Good to see the Sport did so well!!

Yes, those are ID numbers and yes, there is a typo- that Colt is 883.

As I look at the data there seems that there were probably some fliers that unsusually opened up the Colts groups (like the 8" and 5" groups).
But the most interesting thing (beyond the performance of the Sport) is that the data would seem to indicate that the Sport was pretty much settled in right out of the box, while the Colt needed the 1200 rounds to break-in and get even close to the Sports "out of the box" performance.

At the $.32 per round I pay for PCM X-TAC XM193 delivered in bulk, that's about $385 more to settle in the Colt ....... or enough to buy a fairly decent scope, some reloading equip, a really good chronograph, a target trigger and some more mags, an entire CMMG .22LR upper, or take a course from the good folks at LSU!!

But ...... if you take those savings that the Sport has out of the box compared to what it took to get the Colt to settle in, and also add the ~$400 more it cost to buy the Colt, you can buy another complete Sport ......... and have change left for a Happy Meal or two at Mickie-D's!!

So I guess the question becomes ..... can Sport owners be as smug as Prius drivers now?!?!

On Edit: I decided to add the data table here..........

If we throw out the 8+ inch group, both the Colt and the S&W saw their best 8 groups shrink about 35% after firing 1200 rounds. The Sport didn't have its worst group(hornady 75gr) from the first round in its average, because it occured on 7/6/12. As both rifles better group sizes shrank over 30% after getting shot 1200 times, counting the cost of the ammo against only the Colt is fallacious.

Ahh ....... good point!
But heck ...... that cuts my $ punch line in half!!

Interesting results to say the least. Thanks for the update, Matthew!!

"So I guess the question becomes ..... can Sport owners be as smug as Prius drivers now?!?! "

Why, I'll bet if we try real hard we could be as smug as Colt owners

Actually I just did the math and the average of all the rounds that were shot both in the Sport and in the Colt work out this way ..

The Spot averaged 1.753 when new, and the Colt averaged 3.073 when new (both the numbers for the Sport and Colt that had the Colt at 8.311" were tossed).

The average for the same rounds shot yesterday are ...

Sport 1.445, and the Colt (after tossing the 5.295 group) group, was 1.603.

So my initial statement that the Colt took 1200 rounds to get to 'almost' were the Sport started out, is in fact correct ..... unless you are saying that there is missing data that would show the Colt settled in sooner than the 1200 rounds, or that you are throwing out 'all' the Colt's large groups.
You can certainly do that ......... but it's called "cherry picking".

All of the accuracy data gathered from when we put the 3.5-10x leupold and shoot them at 100 yards for groups has been and will be published.

We do not know at what point either gun "settled down" or even what caused it, including external factors, but by all means, speculate.... We may get ideas about what we can control for down the road in order to develop a better data set.

Please do not be smug. I know that when I am smug, something turns into a train wreck. If Sport 095 isn't chugging along 20,000 rounds from now, Groups less than 1/2 an inch smaller fired yesterday ain't going to mean squat.

Reviewing the ambient condition data, the fact that it was 92 degrees and sunny on June 22, 2012 and 73 degrees and overcast yesterday, yesterday's abscence of mirage may account for part of the difference from the first group to the second. If the groups open back up on a hot, sunny day, that would be referred to as a clue!

One of the points that this chart does demonstrate is that to get any real idea of what the actual ability is, you need more than 5 shot groups to see the true performance of a firearm.
There are just too many things that can effect such a small sample in any population.

Although for zeroing in it might be fine .... I'm sure a 20 shot per type of round, per target (irrsepective of how impratical that might have been) for each gun, probably would have yielded quite different table numbers, and (I believe) support your position that the Colt costing more to break in ...... is not correct because the 'largest numbers' for the Colt are coincidently erroneous.

Although it would still demonstrate that the Sport did better than the Colt, it just doesn't make sense that the Colt did both as 'comparatively' good as it did 'and' as bad on the same day, unless 'possibly' it had something like a bimodal issue with the forearm.
I.e. possibly a burr in the hand guards that caused the barrel to rest at two different places, and after firing it a number of times, wore away and finally allowed the barrel to rest at the same spot each time. (?????)

Had I seen that large of differences on day one, I might have popped off the hand guards to see if that might have made it better.

But, seeing that the table's numbers are what they are ......... and this 'is' an M&P forum, we ain't gonna be so giving ..........

Flyer91,

You are the only person who has in any way implied that the Colt cost more to break in, so you are now debating with yourself.

We picked 5 shot groups because they can be easily duplicated by anyone who wishes to buy a box of ammo, shoot a group or 2 for comparison, and have some left to hunt with, shoot for fun, or whatever. We want to produce data in a process that anyone can replicate to determine how useful the data is for them. As this test will go on for years, before it is over, we will have enough 5 shot groups for the data to be meaningful. The test isn't going to be over until one of the rifles cannot be repaired.

Lastly, nearly everyone knows that free floating a barrel will help accuracy. Modifying the rifles to make them more accurate or dependable defeats the purpose of this test. If I put a BCM BCG in the Sport, wouldn't some folks cry foul?

I love the table....and am going to use it everytime someone throws "the chart" out there!!!!!heck it deserves its own sticky. I would also wager that both rifles will be chugging along 18,800 rounds from now! But that will tell if the engineering data on melonite is as stout as they claim!!

__________________
Grover Oathkeeper

Whoa there Matthew, I'm not saying you guys did anything wrong or misguided.
I'm saying that I 'agree with you' that the first day data makes the Colt look worse than it probably was, and I'm not debating the "price to break it in" with anyone ....... I'm just talking like we we're sitting at the range discussing the data, and am trying to get a better idea of what it means.
In other words my comments are evolutionary ........ and admittedly are probably something the neither Politicians. Military Officers, lawyers and internet posters should normally do.

Any aggregate of many samples will always sort out the facts much better than a smaller sample. But ....... sometimes that is not practical nor targeted for whatever reasons, and I guess I failed to make 'that part' of my comment about the 5 rounds clear.

I'm also not dissing you for not looking at the hand guard ..... or am I even saying it was the issue or suggesting you guys didn't know that, just offering some off the cuff reason (other than you flinched a lot more when firing the Colt, and were totally off your caffeine when firing the Sport ) for the data that the Colt evidently produced.

Something obviously (and I suspect 'uncharacteristically') happened to/with the Colt to cause the huge swings in groups from >8" to >3" then to more in line with the Sport at <2", on the first day, that didn't happen to the Sport on the first day nor to the Colt on the last day ......... according to the data.

IMHO, having to do something like removing a burr from a plastic hand guard doesn't necessarily speak ill about one or the other product, but in all honesty ....... I wouldn't compare that to swapping the stock BCG to something like a BCM product, and yup ....... I understand that it's not rocket science.

At first glance at the data I did think it was the needed break-in for the Colt. I mean ..... the data does sort of look that way.
But after giving it some more thought, I posted that 'that' doesn't make sense (agreeing with you) because the Colt couldn't be both that bad 'and' that much better at the same time. I'm sorry I didn't make that more clear.

In retrospect, I probably shouldn't have added the joke about the forum (being M&P folk) not letting go of the Colt's fliers off the hook, and getting to be smug.

But heck no Mathew .......... I appreciate you guys taking the time and trouble to do this and provide us the information.
Once again my sincere apologies if I offended you, it was certainly not my aim.

Flyer brought up a point I had been considering all the way down the post, which is the human element. How did you guys attempt to limit it as much as possible?

No offense taken at all. Your posts have caused me to think about the process, the variables, the data, everything really.

Before we started discussing things, I had not even considered things like mirage. From now on, we can evaluate mirage and note it with the data. One more thing that can help us better understand the data.

At this moment, most of the details of yesterdays activities are fresh in my mind and I cannot possibly record it all. Vigorously discussing the data and how it was collected will help me record and manage variables. It will also make me a better rifleman and a better trainer.

What! No TulAmmo!?

I went back and looked at my comment in post #15 saying ...... " Although for zeroing in it might be fine .... I'm sure a 20 shot per type of round, per target (irrsepective of how impratical that might have been) for each gun, probably would have yielded quite different table numbers, and (I believe) support your position that the Colt costing more to break in ...... is not correct because the 'largest numbers' for the Colt are coincidently erroneous."

Should have read ........... "and (I believe) support your position that my deducing that the Colt costing more to break in ...... is not correct because the 'largest numbers' for the Colt are coincidently erroneous."

I think if you re-read the original comment in the context of what that paragraph is saying in 'it's entierty' .......it really 'does' originally say that, but admittedly not as well as I had hoped.

We prepped everthing. Targets were on target boards, cartridges were in magazines, shooting bench was set up, data pad set up, everything was ready to go so the groups could be fired without interuption. Forestocks were in a Caldwell rest, with a gravel bag under the butt stock. Shooter was rested and ready and took a 30 minute break between shooting the Colt and the Sport. Groups for each rifle were fired in the same order so if fatigue played a role, it affected both rifles respective groups equally. Shooters history suggests human error should be .2 - .5 moa 95% of the time.

That makes more sense to me. Thanks for clarifying.

You wouldn't happen to be able to test the M&P15T vs. SP6940, would you? Thanks to the folks at LSU?

We have some, along with Georgia Arms, Federal xm193, Federal 55 gr .223 fmj, Corbon JSP, and a few other things, but adding more ammo types complicates things exponentially, not linearly like one would expect.

I anticipate that once we ID loads that the rifles really don't like, we will rotate other loads into the testing. After all, in real life who seeks to shoot inaccuarate loads. We think a better comparison is looking at what the Sport shoots really well next to what the Colt shoots really well.

If you send them to us, we will shoot them and test them, but we do not have any more rifle buying money in this years budget. We do have a contingency fund that might get utilized if events on November 6, 2012 indicate that economic, social, and political stress levels will remain elevated. If that congency fund gets tapped, we may have to buy what is readily available.

Before we leave the subject of accuracy testing.

As stated earlier 'if' you can ........ it's much better to shoot something like 20 rounds to get a really good idea of what a weapon is capable of.
But sometimes that isn't desirable because (along with other reasons) it creates too many hits on a target, and you loose your point of reference ........ "was that hit the one I just shot or the one, or two, or ? before that??" sort of thing.

But what you can do is shoot 5 rounds per target into 4 different targets, and then after the shooting ..... place one target at a time over a fresh target (or even a blank piece of paper) and mark the hits for each, on that fresh target/paper.
The marked target can many times tell a hidden story that you totally don't see on the individual 5 hit targets.

Also ......... 20 rounds becomes a sort of "force multiplier" number in shooting and determining what's happening, especially useful for working up a hand load, and doubly so if also using the data from a chronograph.
And because most boxes of ammo conveniently come in 20 round boxes, as the statisticians here know, 20 becomes the threshold for a good sample size statistically.

We can use a box of 20 rounds to make 4 groups of 5 shots per target. calculate the average spread for each target, and then calc the average of those 4 averages (X bar) and to use that to determine the standard deviation.

Although a bit boring .......... 'standard deviation' is exactly that, the amount any given (in our case) single hit deviates from the average of the averages of all the hits, irrespective of which target it's on, when it was shot, or who shot it, and for us ....... that provides a base to determine any change/drift/variations in a weapon, shooters, or ammo, etc., over time.

Now ....... this is probably even 'more' boring ....
Once you have established X bar as above, you can then plot each subsequent single hit on an X bar graph.
Since X bar is the absolute average of the averages and is a value the represents the "accuracy circle" that our system of weapon, shooter, and ammo shoots ..... on average ..... any subsequent shots from then on should fall evenly distributed above and below the value of X bar (IOW: they hit in equal numbers either outside or inside the value that represents that 'best' circle that the weapon, shooter, and ammo produced, by hitting closer or farther away from the point of aim, and when plotted they are equal in number of dots both above and below X bar).
If they aren't something has 'statistically' changed.
If you see a run of 7 shots in a row plotted on the chart that are ascending or descending ........ something had changed.
Keep in mind that chage can be just the shooter on an off day and be a flier(s).
But if the above trends continues it is a certainty that it is not the shooter (unless of course, you are plotting over the shooters initial learning curve ...... or entire lifetime).

You can also further refine the X bar chart by applying statistical upper and lower control limits.
Then the X bar chart becomes a contol chart, and if you see 3 shots in a row either above or below the upper or lower control limit lines ....... something has changed.
But that is REALLY flippin boring, and I've probably already put you guys to sleep ...........

OK ........ so you might not want to get this deep into statistical analysis of what's happening, or to this degree ..... and I certainly don't blame you.
But don't throw your 5 shot targets away.
Instead simply do as the second paragraph above describes and I believe you will see a much better story about what you, you gun and ammo are really doing.
If you see that most of the hits are grouping tighter than any of the single targets show, you could probably assume the weapon is better than a single target demonstrate, and you might just need more practice, and/or your ammo is not consistent, and/or the environment changed .... etc.
If you see there are two (or more) separate tight groups on the combined 'marked target', you might assume something like you shifted your mount on the weapon, or shifted your natural point of aim, or possibly the barrel fouled, cooled, heated ... etc.

It would be interesting to do this and see what multiple 5 shot group targets, using different ammo, overlaid ..... would yield.
Any tight groups on the marked target would certainly be something to ponder.

One of the interesting things to me was how different loads of the same weight and bullet type shot to widely different points on the target. For example, the center of the AE 50 gr Tipped Varmint group was just over 4 inches to the left of the center of the Fiocchi 50 gr v-max. Different ammo always shoots to different points of impact, but I usually see more dispersion in the vertical than the horizontal from group to group. Also, the greatest dispersion is usually between bullets of different weight. We have never tested 11 different loads through 2 rifles before either?

As a side note, the Burris mount returned the Leupold scope to within 2 moa of point of aim when we remounted the scope on the Colt.

The reason we are publishing the raw data, and ALL of the raw data, is so that anyone can apply whatever means of analysis they choose. If you would like us to collect additional data on subsequent tests, let us know. I find it preferable to minimize trips downrange because the time that takes allows environmental conditions to change and lengthens the amount of times the shooter needs be focused and increases the number of times he needs to get focused, which in the end widens the scope of influence of external variables. We can easily set up targets any way we want to before going hot.

Time to plot a scatter diagram ........

Boring to be sure, but can really help sort out that type of thing.

Matthew, I have noticed the verticle dispersions on different types of ammo and have attributed that to my shooting, But it does make you wonder, you would expect a horizontal difference but the verticle dispersions would seem to infer a corkscrew effect in the ammo maybe due to the bullets not being as balanced and as they go down range the corkscrewing would get bigger. I have played around with some tracers just basically seeing if I could lob them onto a spot 800 to 1000 yards down range and never noticed the cork screwing but maybe it is too small to see with the tracer. Using kentucky windage I came real close but I was just screwing around!

__________________
Grover Oathkeeper

I ain't near as smart as you guys, but this is what I've 'figured out'. Some guns are more accurate than others, some ammo is more accurate than others, some guns shoot some ammo more accurately than others. I prefer guns and ammo that put bullets where I want 'em to go.
Gettin WAY too complicated for a country boy.

A cursory look at your chart seems to show me that the Colt prefers the heavier grain bullets, while the Sport favors the lighter.

Nice data on the rifles - Thanks


Sent from my iPad using Tapatalk HD

I am confused by this post.

I would think the verticle dispersion is very common and easily explained, because of different velocities in a variety of ammo, and horizontal dispersion from ammo would be less common in different brands using the same bullet type and weight.

You are so right....I have it backwards.....I had major brain gas when I was thinking of this....I had a bad day yesterday and it shows tremendously!! OOPS

__________________
Grover Oathkeeper

I completely understand.

I thought one of us must of had something backwards.

I am very North and South oriented, but I get East and West backwards all the time.

So how long between shots and shot groups? Just curious to see the characteristic of change with the two different barrel materials as they heat up like in a battle situation.

__________________
-Dik

Each 5 shot group was fired in about 10-15 seconds. Unload, reload, and shooter settling back into position takes about a minute. Each 11 group series took about 15 minutes. The rifles fired 10 rounds at 50 ft to 50 yards to zero/confirm zero before firing the series of groups. This was done with the wolf 62 gr fmj. We shot the ammo samples in the same order each time so that the barrels would be at the same point in their heating cycle when firing each group, respectively.

Ok, thank you. I guess I didn't get across my question very well. When shooting for accuracy I know you're supposed to let the barrel cool for best results, but how do the barrels stack up when they get really hot like 30 rds in a minute or so. I just haven't read anywhere if the melonite compares at real high temps as you'd see in high round counts in a situation you'd see our troops in. I apologize if it's not a pertinent question for the testing you are doing. Great work and thanks for all the info tho!

__________________
-Dik

We are not letting the barrels cool at all. We are firing each string of groups as fast as we can aquire a good sight picture and press the trigger. Split times are 2-3 seconds. Firing any faster would not allow any type of good sight picture and even trigger press.

We do not spray and pray, nor do we lay down cover fire. To test something the way you seem to want to within the bounds of acceptable marksmanship would be better done on a rifle with 3 shot burst capability, so one could triple the number of hits per sight picture.

Ok makes sense, I just wasn't understanding correctly. Thank you

__________________
-Dik

Ok, can you explain the entire process of how you got the information on the "Table"?

The more I read bits and pieces of how you got your results, the more I am confused.

Is there a method to your madness?

Maybe I am just dense, but I am not seeing the complete picture, and it doesn't sound very scientific or repeatable.

I am not trying to be an *****, I am just really am trying to understand how you got the results.

Thanks.

First off, I am a fan and owner of the M&P15 – I own an ORC and have read the many satisfied posts re: Sporter. Second, I do not have a background in statistics. I can barely recall my 2 courses in statistical analysis some 20 years in the past. Third, I appreciate the hard work, expense, and effort you have put into this project.

However, my untrained eye sees problems with your testing method. One problem in particular leaps out of the data. Why shoot one set of 55gr, yet you shoot three sets of 50, three sets of 62/63? The 55 gr set is the ONLY in which the Colt is superior after breakin, and 55 gr is generally the cheapest, most widely available .223 ammo. Why did you choose to use one set of data for this bullet weight, yet used 10 other weight bullets, some multiple times for data sets, especially given your “average of eight” score?

Not trying to diminish your work so far, but any time this type of data is posted it creates a lot of questions, as it should.

The Winchester Q3131 55 grain fmj that we shot is m193 just like the Federal m193 that proliferates the market place. Since most of the 55 grain fmj being sold is mil spec m193, testing multiple samples would not tell us as much as testing samples of ammunition made to differing specifications. We are also trying to identify several ammo types that are the most accurate in each rifle, and m193 does not have a history of being very accurate. We tested fed 62gr otm(mil spec SOST mk318 mod 0) and Blackhills 77gr matchking (mil spec mk262 mod 0) because they are specifically designed to be accurate from the Colt, and they were both very accurate. 1.2, 1.3, and 1.6 moa represent impressive accuracy in any rifle, and from a chrome lined barrel that isn't free floated, it is terrific!

Lastly, we are not choosing what data to use or what data to publish. We are using and publishing all of the data that is generated. If you believe that the ammo that we have tested the rifles with is somehow biased in one rifle's favor, send us some ammo to test and we will publish the results. The samples first tested on 7/6/12 were added at the behest of a reader and we are more that happy to indulge reasonable requests. We happen to be in a position to collect and publish some data that people may be interested in. We will leave it up to rifle and ammo comsumers to analyze the data and decide what it means to them.

find it hard to say that the colt is designed for the ammo when the military isnt using the 55gr anymore, it just seems that that particular rifle liked that ammo on that day! just my opinion.

__________________
Grover Oathkeeper

The Colt wasn't designed for that ammo, but the ammo was designed to be fired from the Colt M16A1. It had to meet an accuracy spec and even though m193 isn't widely used by the military any more, neither it nor the rifle has changed significantly. The mk 318 and mk 262 were designed to be accurate from both the M16A3 and the M4. The factors affecting ammo compatability include chamber and throat dimensions, barrel twist rate, and barrel friction.

My apologies I misread something and got it confused, we had a death in the family and I wasnt thinking to clearly. But I can see why it is better to not have a brand new rifle issued to you in the Military. But if the lining of the barrel changes that much of a short period of time, does the chrome lining continue to wear at that rate or is that something we are going to see after more rounds. Yeah somewhere I got it in my head that the M16 was designed for 55gr fmj and just screwed the hole posting up!!! LOL

__________________
Grover Oathkeeper

This is not too difficult to explain for folks.
The 2 barrels are very different with respective coatings. The melonite treeatment affects accuracy less than chroming a barrel does. And, as a treatment, it does not alter the lands much.
Hard chroming the barrel adds material, and it definitely alters the seating of the bullets on the lands, and in turn affects accuracy. And, as shwon....... the barrel will need to really be 'shot in' with quite a few rounds.
It also shows something else: You are NOT going to hurt a barrel/ accuracy with a wire bore brush. Unless you ram that brush thru the barrel at about 3000fps about 100,000 times.
This should also help shooters understand that when not using hot loads, your rifle barrel WILL last a lifetime of shooting. Whether the barrel is chrome lined, melonited, or just plain 'ole hard steel barrels. My sport is now on the 10,000 round mark, and accuracy is still equal to the day I started shooting it.
For those who would wonder what it takes to get a REALLY accurate, jewel quality chrome lined barrel? The barrel is rifled slightly oversized. Then, the barrel is chromed to a double thickness. Third and final step, the barrel is then re-rifled. The new cuts will only be in the chrome, and the sizing will be true to the spec for the bullet out of the box, with no flaws in the edges of the lands& grooves. It does cost some extra money for this extra work.

Looks like the Wolf ammo seems to stack up pretty well from the chart. This mirrors my own observations with it.
Great job on the chart and ammo- a real nice spread of rounds chosen!!!

Once the barrel is broken in, the accuracy should stay remarkably consistent for between 25-50,000 rounds. In the break in, all edges that have microscopic flaws and surface irregularities need to be polished, and the only way to do that is by sending ammo down them. Once seated, the rifle meets a 'comfort zone' for pressures, friction, etc.
Analogy: Breaking in a new motor- rings need to seat in the cylinder walls. If you have ever seen a cylinder that is fresh?? It's actually a bit rough, with cross-hatch in the cylinder rather than a perfectly honed unit. But after a couple high rpm runs, the cylinder is smooth as butter, and compression hits its' real numbers. (This grinding in is why you change the oil after say 500 miles the first time, and DO NOT USE synthetic oil for the first running of the engine. Synthetics do too good a job, and cylinders would not seal completely.

Good question!!

Can you tell us who makes these barrels, and where they are available?

Thanks.

Rojo, can you cite prior research that parallels re barrel wear and break in? We are always interested building on and comparing the work of others.