The Not So Masterful Thief: Prop Lockers Are Lame

[LaznAzn]

So I have recently got my hands on my first Master Thief mod and the first thing that came to mind was "Hey, what about those pesky Corpus locker rooms that are almost always full of red lockers!" Being as sneaky as only a Tenno can, I happened upon one of these locker rooms while farming neural sensors from Themisto. I was greatly dismayed to find I barely opened any of them at all! I wanted to blame the mod but it had served me well prior to this (at the expected 40%) so I decided it was simply bad luck.

 

Then I popped into Outer Terminus for a little relaxed mod farming, as we all know these locker rooms are a common starting location for these particular line of defence missions. Again the open rate was abysmal. My frustration motivated me to run a little experiment that I will now post below:

 

 

This is a visual representation of a top-down view of a typical Corpus locker room, as you can see there is a whopping 44 lockers in this room. I'm sure many of you thought it a great place to use your Master Thief mod, unfortunately you will see it is only a little bit better than other locations.

 

For this experiment: Red X'd squares will stand for attempted but still locked lockers, Dark Green squares will represent lockers that were already open, Blue squares represent lockers that were unlocked by the Master Thief mod. (maxed to a 40% success rate)

 

All runs took place on Venus-Kiliken. Some may point out that this might not be the case everywhere, there is certainly that possibility but I will at least demonstrate an issue with this location and it is up to you to decide whether or not it applies to all similar locations or not.

 

 

This was my first test run, maybe I was unlucky? 

 

 

Hmm, second run was pretty bad too...

 

 

 

 

 

 

 

 

 

And there you have it. Either my luck is the worst in the world or there is something about 31 of the 44 lockers in this location causing them never to open. My idea is that they are static props that are never unlocked.

 

 

The above diagram shows you what lockers in this room can be opened either naturally or using Master Thief. Now I'm not saying this should be changed or that Master Thief is useless. When farming for resources every little bit can help and Master Thief can certainly net you more than a little and I will continue to use it regardless if this "problem", if it is a problem, is addressed or not.

 

I'm sure the images speak for themselves and the final diagram gives a clear conclusion to this post. I'm just trying to say Master Thief opens locked lockers, and I had it in my mind "red lockers" were "locked lockers" and thus can have the possibility of being opened. Then again, maybe I am the unluckiest person in the world!

 

Good day to you all folks, and thank you DE for such a great game to spend some of my summer break 

 

EDIT: Improved images for our colorblind Tenno comrades! Also, over these 10 runs, if you assume the the red lockers in the last image are all props, there's ~25% of the 13 lockers opened naturally and ~25% of the 13 lockers opened using Master Thief.

Edited June 3, 2013 by LaznAzn

[MetalGerbil]

the problem with statistics is that you need thousands of samples to get a good representation.

good work none the less. wish i had my own to test.

[.PiTHON.]

the problem with statistics is that you need thousands of samples to get a good representation.

good work none the less. wish i had my own to test.

 

Not when the chance is 40%. The game is clearly coded to have some lockers that are "real" and it likely randomly chooses to have these real ones be green and openable or red and locked, while the rest are just "prop" lockers as OP called them.

[TunaMayo]

I believe you may be onto something with the static props, these are some of the very first tiles made so wouldnt be surprising. Also now I think about it there are certain lockers in those type rooms I have never seen open, ie in the corner.

Also it's statistically improbable that it would be the same locker 5 times to open, and the others in the row aren't, cba to do the math.

Would like DE to acknowledge and state if these are props or not, would save people hours of testing.

[yohaseo]

i think there may be a better chance of unlocking the others by having other ppl in a group with it as well

[TunaMayo]

i think there may be a better chance of unlocking the others by having other ppl in a group with it as well

I doubt this very much as it is stated as a flat 40%.

One sure fore way of testing this would be to get a group of four people with the mod and run up to each container one after the other. If none of the "static" boxes open you pretty much have your answer.

[winterau]

the problem with statistics is that you need thousands of samples to get a good representation.

good work none the less. wish i had my own to test.

 

This is simply not the case with this particular test.

 

If you have a 40% chance (0.4) to open any single red locker and a locker is red in 10 tests and you never open that locker, you chance of not opening that locker (.6) in 10 attempts is 0.6% (0.006). 

 

Now given that you have a whole bunch of red lockers that stay red through 10 tests, I would be comfortable saying with a very high degree of confidence that some lockers always stay red and will stay red even with Master Thief active because they are not coded to ever be openable.

[Worira]

Considering that I'm pretty sure this room is one of the main reasons people wanted Master Thief, this is pretty silly.

[TunaMayo]

This is simply not the case with this particular test.

 

If you have a 40% chance (0.4) to open any single red locker and a locker is red in 10 tests and you never open that locker, you chance of not opening that locker (.6) in 10 attempts is 0.6% (0.006). 

 

Now given that you have a whole bunch of red lockers that stay red through 10 tests, I would be comfortable saying with a very high degree of confidence that some lockers always stay red and will stay red even with Master Thief active because they are not coded to ever be openable.

Get four people at once with the maxed mod on. 0.6^4 is ~13%, that means if you are running with four of these maxed mods it would be interesting to test and theoretically, a container has a 87% chance to open.

But it wouldnt necessarily show that they are props, it could be that it is pre-determined at the beginning of the mission just as one person has the mod on.

Or is it that if you use the mod and it fails then it perma-locks? If so thats a dhame. Would be nice to have a team that opens 87% of lockers.

[winterau]

Get four people at once with the maxed mod on. 0.6^4 is ~13%, that means if you are running with four of these maxed mods it would be interesting to test and theoretically, a container has a 87% chance to open.

But it wouldnt necessarily show that they are props, it could be that it is pre-determined at the beginning of the mission just as one person has the mod on.

Or is it that if you use the mod and it fails then it perma-locks? If so thats a dhame. Would be nice to have a team that opens 87% of lockers.

 

I am pretty much certain that they are props.

 

The prop hypothesis for a particular locker can be disproven by the locker opening in one of the 10 runs. The chance of that not happening is  (0.6^10 = 0.006) i.e. 0.6%.

 

There are 31 lockers that don't open. The chance of at least 1 of those 31 lockers not opening in 10 runs is an astronomically small number.

 

They are props, I would bet my house on it.

[Artaban]

The 40% could be like crit mods

 

Say each locker has a 10% of being unlocked in a given mission, and are generated as the mission loads. Equiping master thief allows you to reroll lockers with a 40% chance greater than baseline, i.e. 14% of being unlocked.

[Zepheros]

The 40% could be like crit mods

 

Say each locker has a 10% of being unlocked in a given mission, and are generated as the mission loads. Equiping master thief allows you to reroll lockers with a 40% chance greater than baseline, i.e. 14% of being unlocked.

 

This actually makes more sense.

[kaizoku222]

the problem with statistics is that you need thousands of samples to get a good representation.

good work none the less. wish i had my own to test.

Entirely wrong.

 

In order to get margine of error down under 10% you only need 30 trials, and the sample size the OP has is sufficient enough to make a weak conclusion. It would only require about 19 more trials to *solidly* conclude that that specific set of lockers has a massive amount of "dud" lockers, and again even 11 trials is a good start.

 

If the mod doesn't function like it says (ala crit mod), then it's a problem with the text and it's DE's responsibility to fix it to state what it does accurately. Waaaaaay too many things in this game don't actually do what the flavor text says, and considering that precedent, it's very likely DE has messed up the description for this mod.

[Volt_Cruelerz]

OP: I really would like to follow along, but I'm sorry, for the colorblind, your post is very difficult to find what you mean.  Could you please make the green a darker shade or something? :/

[matrixEXO]

Off-topic question: How and where did you get the Master Thief mod?

[Onite]

Not when the chance is 40%. The game is clearly coded to have some lockers that are "real" and it likely randomly chooses to have these real ones be green and openable or red and locked, while the rest are just "prop" lockers as OP called them.

Yeah i noticed this too and its been pissing me off greatly. Most of your resources come from containers and lockers. Through the patches leading up from 7 to 8, i kept noticing that DE was taking out more and more lockers from certain tiles all around on every single tileset. And the number of locked to unlocked lockers also dropped considerably. So this thread really doesnt surprise me and only further highlights this. If i didnt know any better, i'd say DE is trying to curb your resource gain rates as much as possible to stretch their content. Which also explains why they would also move almost all the weapons over to blueprints now instead of credits.

 

BullS#&$ move, DE. BullS#&$ move.

[TheClingClang]

Aren't we all happy the design council picked this over something cool like pocket rockets?

It was a S#&$ty idea, so DE made a predictably S#&$ty mod. I see no problems here.

 

Also wow is there a lot of gambler's fallacy going on in here. You're also assuming that the percentage relates directly to the chance that a given locker will open, and not to some equation like Streamline's power efficiency.

Edited June 3, 2013 by TheClingClang

[Worira]

Aren't we all happy the design council picked this over something cool like pocket rockets?

It was a S#&$ty idea, so DE made a predictably S#&$ty mod. I see no problems here.

 

Also wow is there a lot of gambler's fallacy going on in here. You're also assuming that the percentage relates directly to the chance that a given locker will open, and not to some equation like Streamline's power efficiency.

 

This is, uh, not the gambler's fallacy. At all. It's basic statistics.

[winterau]

Aren't we all happy the design council picked this over something cool like pocket rockets?

It was a S#&$ty idea, so DE made a predictably S#&$ty mod. I see no problems here.

 

Also wow is there a lot of gambler's fallacy going on in here. You're also assuming that the percentage relates directly to the chance that a given locker will open, and not to some equation like Streamline's power efficiency.

 

You have no idea what gambler's fallacy is if you think that it is applying in this case. 

 

No one is making any claims about the chance of any given locker opening based on the behaviour of previous lockers. What we are saying is that for a run of 10 times, the chances of the locker not opening in any of the 10 runs is 0.6^10.

[Nehmetawy]

As a curious thought, what was your locker totals for across the entire mission?

How many were Green lockers per mission?

How many were Red?

How many became unlockable?

[SymphNo9]

If some of the lockers are indeed props, then this is disappointing news.

[IcarusNine]

I've noticed this, especially in this room.  There are always certain lockers that have never been open.  The 12 lockers on the left, the 'entry way' if you will, is probably the most obvious to see this, since 10 of them never open while the other 2 have comparably high chances of being open.

 

By the way, the right-most locker of the OP's first run isn't shown as being a 'real' locker in the final diagram.  Judging by the OP's estimated 'real' lockers and that one extra, there are 14 'real' lockers.  The OP encounters 32 green lockers out of a possible 140 lockers, which suggests a  22% chance of starting open.  One could estimate the chance being 20% or 25% (since devs are shown to like increments of 5%).  That yields a 45-48% chance (0.6*0.75 or 0.6*0.8) that any particular locker will both start out locked and refuse to be opened by Master Thief.  Over ten runs and 44 lockers, to see that 30 lockers never open would be astronomically unlikely.

 

 

Of course, you just know that DE's solution to this problem will simply be to remove the inanimate lockers.

[Law_Abiding_Engineer]

Question, did the lockers that were "locked" but unlocked thanks to your mod, have a different color to them, or were they the exact same shade of red as the lockers that were actually locked?  I mean, thats kinda a bummer because you have to go to all red lockers and try to open them, and if you are on a crappy server with lag, then you have to wait a while in order for the locker to open.

[LaznAzn]

OP: I really would like to follow along, but I'm sorry, for the colorblind, your post is very difficult to find what you mean.  Could you please make the green a darker shade or something? :/

 

I hope the new images will make things a bit easier to understand now! 

 

As a curious thought, what was your locker totals for across the entire mission?

How many were Green lockers per mission?

How many were Red?

How many became unlockable?

 

I wish I could tell you but I wasn't really keeping track of that as I those locker rooms appear right at the spawn so i could abandon out to do another trial without need to actual explore further. But I do plan on keeping a count on how many lockers I encounter from now on and how many were naturally unlocked or required the mod to unlock.

 

Question, did the lockers that were "locked" but unlocked thanks to your mod, have a different color to them, or were they the exact same shade of red as the lockers that were actually locked?  I mean, thats kinda a bummer because you have to go to all red lockers and try to open them, and if you are on a crappy server with lag, then you have to wait a while in order for the locker to open.

 

No, when you approach a locker it will automatically either turn green and become unlocked or remain red. There is no need to press the interact key, the whole process is automated. Well, not exactly true, when a locker turns green you will have to open it just like any other green locker. 

Edited June 3, 2013 by LaznAzn

[.PiTHON.]

No, when you approach a locker it will automatically either turn green and become unlocked or remain red. There is no need to press the interact key, the whole process is automated. Well, not exactly true, when a locker turns green you will have to open it just like any other green locker. 

 

Oh that's kinda cool. I figured you had to walk up to each one and try to open it, and that it would still be red even though it opened. It's nice that you can just walk by them instead.

 

And although I think it's lame that there are prop lockers, I don't think people should be so quick to call DE a S#&$ty company and that this is the fault of S#&$ty idea from the S#&$ty design council. Calm down. For all we know load times were being affected from having to process rooms full of real lockers so they saved some CPU time and made most of them props.