Mantis Training Device

[nycnoob]

INTRODUCING MANTIS TRAINING DEVICE

http://www.guns.com/2015/09/01/mantis-pistol-attachment-and-training-to-improve-precision/


http://mantisx.com/


Once the Mantis is attached to a handgun, it keeps track of the barrel movements during the trigger pull and, from that data, it determines which direction from the aiming position the gun moved and by how much. This information is then displayed on the screen of the phone: the direction on the wheel and the magnitude of the movement through the score (100 equals no movement).




MANTIS
$ 169.99 (normally $ 199.99)
Pre-order now and save $20 with code MANTIS20!



3839


Saw this device on guns.com site today. It looks really interesting however I wonder if it is actually useful. Does one really need to understand the cause of their errors in order to fix them. I want to improve my shooting not necessarily understand in the minutest detail what I am doing wrong. In other words perhaps correct dry fire will teach you good habits without you ever consciously knowing what the problem was previously.

[BaiHu]

Intriguing...

[Peally]

I don't know how useful it would be but it'd be interesting to know which direction I'm ganking the gun on particular shots. Regardless good trigger control will iron out any issue.

[45dotACP]

Why not just use a laser?

Sent from my VS876 using Tapatalk

[Alpha Sierra]

Does one really need to understand the cause of their errors in order to fix them.

Uhhh, yes. Anything less is guessing.

I find it hard to believe that anyone would question that......

[Alpha Sierra]

A more sophisticated option: http://www.scattusa.com/

[Clusterfrack]

Very interesting. I think it uses an 3-axis accelerometer to detect movement. It's disappointing that "At this point, it can’t detect double taps or super rapid firing, but it can track consistent rhythmic, static shots." That might be a limitation of Bluetooth bandwidth not supporting a high enough sampling frequency.

[45dotACP]

There are better options I think...specifically reading your sights and calling shots. It's not as crazy difficult as everyone makes it to be and I think this gizmo is a gimmick. It's probably a kinda strong opinion to take, but I don't think this thing will catch on. It'd be pretty cool if it did, but I'm already at the point where dry fire and some focused live fire drills have essentially brought my slow fire shooting skills to where they need to be.

Also a thing to consider is whether the gizmo, when able to read rapid fire strings...will label your post ignition push as "heeling" the gun because it was only .20 seconds before your next shot?

[orionz06]

Is the target no longer an indicator of movement?

[Luke]

Is the target no longer an indicator of movement?

Just because you hit low left doesn't mean you hit low left. Just buy it.

[Luke]

So if PF buys one, would any of our Site Supporters will be interested in testing it for a week or so?

I'm game!

And I could use the help haha

[BaiHu]

So if PF buys one, would any of our Site Supporters will be interested in testing it for a week or so?
I'm in for a test week and a few hundred rounds.

[scw2]

So if PF buys one, would any of our Site Supporters will be interested in testing it for a week or so?

Probably would be interested. I've been fixing my stance and grip and seem to have improved my left-drift and also recoil management, but I'm still shooting low. I think this would be helpful in addition to the videos I'm watching of myself press the trigger to see what changes to my grip are doing, in addition to just looking at the sights when I press.

[Chance]

Does it work with dry fire? I might could see that being useful.

[MantisX]

Saw this device on guns.com site today. It looks really interesting however I wonder if it is actually useful. Does one really need to understand the cause of their errors in order to fix them. I want to improve my shooting not necessarily understand in the minutest detail what I am doing wrong. In other words perhaps correct dry fire will teach you good habits without you ever consciously knowing what the problem was previously.

Hey guys, Austin from Mantis here, following up on some old questions.

MantisX provides movement feedback, but often the shooter is unaware of what caused it. The system identifies common causes and coaching to help the shooter know what to focus on. Additionally, it quantifies each issue and prioritizes so you know the one thing to focus on (rather than trying to think about 5 different things ineffectively). Once an issue is fixed (quantitatively), it highlights the next most impactful area of focus.

Also, if a shooter is overcompensating or has sight alignment/sight picture issues, etc. looking at a target won't give you the raw data to give an accurate diagnosis of what errors there are.

MantisX handles both live fire and dry fire, and we recommend as much live fire training as possible. Physiological state is much different and has a significant impact on shooting mechanics.



Very interesting. I think it uses an 3-axis accelerometer to detect movement. It's disappointing that "At this point, it can’t detect double taps or super rapid firing, but it can track consistent rhythmic, static shots." That might be a limitation of Bluetooth bandwidth not supporting a high enough sampling frequency.

It has accelerometers and gyroscopes. It does detect double taps/rapid fire, but the issue isn't the bandwidth in this case - it's more in the data analysis: was a movement due to gun movement or user movement? It's a tough technical problem to distinguish, so the app automatically discards data it deems as confusing or unusable in the overall analysis. We will be releasing a future version that accommodates rapid fire scenarios (doing a lot of neural network based machine learning right now as part of that).



Is the target no longer an indicator of movement?


There are lots of external factors that make a target analysis incomplete. Additionally, MantisX works on whatever target, paper or otherwise, multi-target setups, etc.

Holler if you have more questions I can answer.

Cheers,
Austin

[45dotACP]

Not sure if this was already asked, but what's the cutoff time for "rapid fire?"

2 seconds between shots? 5 seconds?

Sent from my VS876 using Tapatalk

[Trooper224]

What's next, an electronic gadget to let me know I'm still breathing? It amazes me that some of us learned to shoot proficiently without a plethora of electronic widgets.

[Mr_White]

Before the Mantis Training Device, I had to train all my mantises the old fashioned way.

---

MantisX, I kid, I kid. ;)

[BehindBlueI's]

Before the Mantis Training Device, I had to train all my mantises the old fashioned way.

---

MantisX, I kid, I kid. ;)

Via prayer?

[BehindBlueI's]

The system identifies common causes and coaching to help the shooter know what to focus on. Additionally, it quantifies each issue and prioritizes so you know the one thing to focus on (rather than trying to think about 5 different things ineffectively). Once an issue is fixed (quantitatively), it highlights the next most impactful area of focus.


Can you break it down a bit more for me? Is it personalized, and if so how? Or is it simply an electronic version of this, offering generic advice based on most common issues?

http://pistol-training.com/wp-content/uploads/2008/06/correctionchartright.jpg

How exactly is it "quantified" and how is "prioritization" established?

[CCT125US]

Is the target no longer an indicator of movement?

That was my question as well. Zero the gun. When bullet impacts somewhere other than POA, pause, reflect, correct. Or in simpler terms: What the hell were my sights doing there?

[MantisX]

Not sure if this was already asked, but what's the cutoff time for "rapid fire?"

2 seconds between shots? 5 seconds?

Sent from my VS876 using Tapatalk

Currently, it will give feedback on shots about 0.5-0.6 seconds apart, though YMMV depending on how much other movement you have.

[MantisX]

What's next, an electronic gadget to let me know I'm still breathing? It amazes me that some of us learned to shoot proficiently without a plethora of electronic widgets.

Agreed, there is a lot of crap on the market. :)

What amazes me is how many people still struggle to shoot proficiently. Before launch, we tested on over 1000 law enforcement, and I'd say a good 20% struggle with some basic proficiency. They have good safety mechanics, but poor shooting mechanics.

[MantisX]

Before the Mantis Training Device, I had to train all my mantises the old fashioned way.

---

MantisX, I kid, I kid. ;)

I hope you were training this kind of Mantis:

https://www.youtube.com/watch?v=F5FEj9U-CJM

[MantisX]

Can you break it down a bit more for me? Is it personalized, and if so how? Or is it simply an electronic version of this, offering generic advice based on most common issues?

How exactly is it "quantified" and how is "prioritization" established?

Good questions.

There is some correlation to the diagnostic target, and we chose the diagnostic target as a primary interface screen because of it's familiarity. And yes, the coaching advice is based off of what the most common issues are.

MantisX goes beyond that in several ways:
1) You don't need a diagnostic target, and can use anything as a target (as well as in scenario-based training), since it is tracking movement.
2) It strips out external issues to just look at raw mechanics. For instance, let's say sight picture is a little off, and they are breaking their wrist down. The shot would land low left or low right, leading you to think they might be slapping the trigger or tightening their grip, misdiagnosing the root problem. Same thing with sight alignment. Or with natural arc of movement, etc.
3) It works with dry fire.

Quantification:
Each shot is scored, based on what we calculate as their deviation from sighted position. There is also an average score for a shooting session.

Prioritization:
Most shooters have multiple issues going on. Rather than telling them 5 things to think about, the system calculates the issue causing the greatest impact to their score (and thus precision), and tells them to focus on that one thing. For instance, simplistically let's say in an 10 round session they were slapping the trigger on 3 shots (with an average score of 75), breaking their wrist on 5 shots (with an average score of 95), and tightening their grip on 2 shots (average score 80). The area of focus would be on trigger control, as it is having the greatest numerical impact. Once that issue is corrected (as calculated in future rounds), the next area of impact is highlighted as an area of focus.

Hope that helps clarify some stuff.

Cheers,
Austin

[JAD]

Would it be possible for a mod to edit the thread title so that the product is properly spelled? It will search better that way, among other reasons.

[rjohnson4405]

So if PF buys one, would any of our Site Supporters will be interested in testing it for a week or so?

I'd be in, although, you likely have much more capable users. I might provide some more novice feedback.

[BaiHu]

I'd be in, although, you likely have much more capable users. I might provide some more novice feedback.
I'd say anyone who can't run a FAST under 6 seconds is probably a good candidate. That includes me [emoji16]

[BehindBlueI's]

Interesting. Does it allow a draw stroke?
I'd be interested in testing.

[MantisX]

Does it allow a draw stroke?


Yes, draw strokes don't interfere with shot detection/analysis (that's where the machine learning comes in). MantisX looks at the snapshot of movement right before and as the shot is going off, so extra movement, though collected, is essentially discarded from the analysis.

Since it's a rail mount, you'll have to have a holster that accommodates it (some flashlight holsters work). We're working on a magazine floor plate mount so you can use your existing holster, though there obviously will be limitations on how many models we'll accommodate.

[MantisX]

Would a wrist mount work? If so, you'd get compatibility with 100% of guns and 100% of holsters. Just a thought.

That's a good thought, and similar devices in other sports (e.g. Zepp in golf) have a hand/wrist mount version. In our testing, what we found was that the accuracy was much better if the sensor is hard-fixed to the firearm. Too much of the tiny movement escapes if it isn't.

[Mr_White]

Nice mantis shrimp video! I meant the insects though.



2) It strips out external issues to just look at raw mechanics. For instance, let's say sight picture is a little off, and they are breaking their wrist down. The shot would land low left or low right, leading you to think they might be slapping the trigger or tightening their grip, misdiagnosing the root problem. Same thing with sight alignment. Or with natural arc of movement, etc.

This part, I'm not following. I think I am understanding that the MantisX detects the direction/character of tiny gun movements just before and while the gun is firing. But does it tell you why those movements are happening? Like in your example, is the MantisX going to know that someone is breaking their wrist down instead of jerking the trigger? Or is it simply going to indicate that the muzzle was low by some amount when the gun fired?

[Mr_White]

Via prayer?

Nope, just the old carrot and stick. They do tricks, they get bourbon. They don't do tricks, you threaten to beat them with little twigs.

[scw2]

I'd say anyone who can't run a FAST under 6 seconds is probably a good candidate. That includes me [emoji16]

Then I REALLY qualify as a candidate :)

[MantisX]

This part, I'm not following. I think I am understanding that the MantisX detects the direction/character of tiny gun movements just before and while the gun is firing. But does it tell you why those movements are happening? Like in your example, is the MantisX going to know that someone is breaking their wrist down instead of jerking the trigger? Or is it simply going to indicate that the muzzle was low by some amount when the gun fired?

This is where the machine learning piece of the algorithms come in. We've taken large datasets and do pattern matching to diagnose the likeliest causes. It's kind of like image processing, if that makes sense. So in the app, it says if the movement was likely caused by jerking the trigger or breaking their wrist down, based on the matching movement pattern.

[Mr_White]

... in the app, it says if the movement was likely caused by jerking the trigger or breaking their wrist down, based on the matching movement pattern.

So, the app is recognizing a gun movement pattern difference between a shot going to spot X because of jerking the trigger, vs. a shot going to same spot X because of breaking the wrist down? Or is it making more of a general probability guess that a shot to low spot X is most likely from jerking the trigger as opposed to some other less likely thing, based on a general perception of 'how people generally shoot and their most likely problems' that was fed into the machine by the design/programming?

Thank you for answering all these questions we are asking!

[1slow]

Via prayer?
I saw what you did there !

[MantisX]

So, the app is recognizing a gun movement pattern difference between a shot going to spot X because of jerking the trigger, vs. a shot going to same spot X because of breaking the wrist down?

Yes. This. :)

There is a little of the second to cover gaps in datasets, but over time it is less and less as our machine learning chews up more and more data and incorporates those improvements into the algorithm.

[Mr_White]

Yes. This. :)

There is a little of the second to cover gaps in datasets, but over time it is less and less as our machine learning chews up more and more data and incorporates those improvements into the algorithm.

Ok, I'm intrigued. This seems like it has potential.

I think there was some talk upthread, not sure how idle, about people here testing these units. I'm game. Regardless of that, thank you very much for the additional explanations.

[MantisX]

Ok, I'm intrigued. This seems like it has potential.

I think there was some talk upthread, not sure how idle, about people here testing these units. I'm game. Regardless of that, thank you very much for the additional explanations.

Happy to answer questions. We're hoping to get some videos out soon that show it in action - a demo really puts everything together.

[45dotACP]

Okay, now my interest is piqued as well...I'd be interested to see a thread or two to evaluate how this thing shakes out.

Sent from my VS876 using Tapatalk

[Sal Picante]

MantisX - you guys are local to Illinois?

If so, I'd love to buy one and try it out - get/give some feedback, etc.

[LSP972]

It amazes me that some of us learned to shoot proficiently without a plethora of electronic widgets.


I know, huh? How did I ever survive the streetz???

.

[MantisX]

MantisX - you guys are local to Illinois?

If so, I'd love to buy one and try it out - get/give some feedback, etc.

We chose the land of the gun friendly, of course. :)

Our office is in Oswego - where you at?

[GRV]

Interesting. Based on the very little I've skimmed here, and the fact that it doesn't require dealing with Windows, I'd be interested in testing a unit if there's going to be a PF thing going on.

[jc000]

So if PF buys one, would any of our Site Supporters will be interested in testing it for a week or so?

Yes…

[Sal Picante]

We chose the land of the gun friendly, of course. :)

Our office is in Oswego - where you at?

Crystal Lake/McHenry border ... Will PM tomorrow and chat. I'm very intrigued.

[Dallasmed]

If anyone in the DFW area wants to get hands on with one shoot me a PM and we can meet up.

[MantisX]

If anyone is going to SHOT Show, we'll be there. You can swing by and play with a demo.

[Sal Picante]

If anyone is going to SHOT Show, we'll be there. You can swing by and play with a demo.

My buddy, owner of Alpha Range, Herb, is going to be out there - I've asked him to find you guys.

Cheers!

[45dotACP]

I'm in the Plainfield area...once I scrape up a few bucks, I'll see if I can throw one on my only railed gun...a Glock 34

Sent from my VS876 using Tapatalk

[MantisX]

I'm in the Plainfield area...once I scrape up a few bucks, I'll see if I can throw one on my only railed gun...a Glock 34


You're close by! - swing by our office and try out a demo first. PM me if you want to set up a time.

[orionz06]

Anything ever come of this?


Sent from my iPhone using Tapatalk