[Finstersang]

People, not meaning to be disrespectful, but your odds are way off. With ONE dice you indeed have 1/6 odds to glitch, but with 2 it already goes to 2.8%. Remember that to glitch you need MORE THAN HALF of your pool, so both dice need to be 1s with  a 2 dicepool.

Oh, you are right. Always thought it was half the dicepool or more 

[ShadowcatX]

So all you did was nerf all your players while making the game more complicated and rolling more annoying? No upside that I can see. You do you, but I'd walk away from the table at that point, too much "be a fan of the players" in my thinking.

[luizborges]

So all you did was nerf all your players while making the game more complicated and rolling more annoying? No upside that I can see. You do you, but I'd walk away from the table at that point, too much "be a fan of the players" in my thinking.

To Nerf would be to make them weaker. That is not true in any way.
Again, according to the books glitches are complications, not lack of skill. The more skilled they are, the better they handle those complications. One complication every 20ish rolls, is that nerfing??

How can the rolling be more annoying?? It is easier and simpler. If the glitch die is anything other than 1 you don't need to bother counting 1s (something that you have to do in all rolls in RAW). If it 1, you still don't even have to check what is half your pool or anything. Put the 1s beside the hits, if it's more, it glitches. Purely mechanical action that is usually done anyway even in RAW to separate hits from 1s, but here you do that only 1/6 of the time. Think about all the time you will save counting ones 

The hit cancelation, still mechanical. Separate 1s that match the count of hits (if you put them side by side it is visually clear even without separation), from what is left, just take out 1s in pairs and bring a hit with you for every pair.

Oh, this is all if you want to go the purely mechanical way, if you can easily see that you are over by 2 remove a hit, by 4 remove 2 hits. You rarely will be over by more than 4 (for that you need a pile that is quite big already).

Remember that the doc I linked is formal description writing. Is how rules are supposed to be written to encopass all possible scenarios. The hit cancelation is described as turning to a neutral face just so that there is no confusion if you use second chance or push the limit. Do you need to turn them to a neutral face,? Not really, I believe everyone is smart enough to notice that you can just disregard those hits in 99% of time when no other dice will be added and no rerolls will be done, but since those are rules, they must be written in an unambiguous way.

[Kiirnodel]

So all you did was nerf all your players while making the game more complicated and rolling more annoying? No upside that I can see. You do you, but I'd walk away from the table at that point, too much "be a fan of the players" in my thinking.

To Nerf would be to make them weaker. That is not true in any way.
Again, according to the books glitches are complications, not lack of skill. The more skilled they are, the better they handle those complications. One complication every 20ish rolls, is that nerfing??

Except you have a mechanic where glitching can reduce the number of hits. That is explicitly making them weaker. No part of your idea balances that part out.

You are changing glitches from just a complication to complication plus you do worse on the roll.
And it definitely increases the chances of critical glitches...

[luizborges]

The reduce hits mechanic is a way of having occasional criticals. Otherwise if the requirement of no hits for criticals would keep then happening almost never. This has minimal impact on pools, as you can clearly see by its odds. To remove just 1 hit has a very, very small probability. I would not consider that nerfing. It could be nerfing if I used that to cap max hits or to declare arbitrary criticals based on some other criteria. As it is I doubt it would be considered nerfing.

The table of probabilities is there, check it yourself (but you are going to need to make some calculations), I even made sure to include a every probably per remaining hits. Just make sure that you understand that is "probability of glitching with N hits remaining" it is NOT "probability of hit cancelation"!!!

Also, don't forget that you always have Push your Limit and Second Chance to get more hits. You will just endure a critical clitch if you really want to.

Finally, my house rule is just that: a house rule. It is not mandatory to anyone. I'm sharing it here because someone might find it interesting. If someone don't like removing hits, no problem, just be advised that critical hits will still be VERY rare, to the point of being barely noticeable above a 2 dicepool. The regular glitching odds will remain just fine.

EDIT:
Here, I made you some probabilities:
Odds of cancelling 1+ hits: http://anydice.com/program/ea5b
Odds of cancelling 2+ hits: http://anydice.com/program/ea5d
The first includes the second obviously. And most of those would result in a critical (this is the main critical mechanism on the rule). You have to understand that to be able to cancel 2 hits you need a pool of at least 8 dice, of which 6 are 1s. To cancel 1 hit at least 4 dice are needed just for that.

[Spooky]

Luiz, while I appreciate your effort at making statistics work, I think you really need to remember that while stats work on paper, in the real world application, stats often fail to explain what actually happens. As my example, I have watched one of my players doing his first skill roll, with a pool of 12 dice. His first result: 12 1s. After the shock passed, I said reroll that. His second result: 12 1s. This is the point when I said "use my dice". I handed him my dice (which unbeknownst to him, had six cheaters in it, so it was not possible for him to roll any more than 6 1s).   His third roll: 6 1s and 6 2s. That makes 36 minimum results out of 36 dice rolled. He gave up playing games where he had to roll dice, because he broke math when he did.

[luizborges]

@Spooky, Bernoulli would disagree with you (as would every casino owner ever) 

[ShadowcatX]

So all you did was nerf all your players while making the game more complicated and rolling more annoying? No upside that I can see. You do you, but I'd walk away from the table at that point, too much "be a fan of the players" in my thinking.

To Nerf would be to make them weaker. That is not true in any way.
Again, according to the books glitches are complications, not lack of skill. The more skilled they are, the better they handle those complications. One complication every 20ish rolls, is that nerfing??

You have made your characters strictly worse than they would be otherwise by causing penalties to come up more often. Yes that is nerfing them.

How can the rolling be more annoying?? It is easier and simpler. If the glitch die is anything other than 1 you don't need to bother counting 1s (something that you have to do in all rolls in RAW). If it 1, you still don't even have to check what is half your pool or anything. Put the 1s beside the hits, if it's more, it glitches. Purely mechanical action that is usually done anyway even in RAW to separate hits from 1s, but here you do that only 1/6 of the time. Think about all the time you will save counting ones 

Ok, I have 16 dice, roll 16 dice, shit I forgot the glitch di.

I have 16 dice, roll 17 dice, shit I rolled all the same color.

I rolled a medium number of ones I have to count. Now I have to count successes. Now I have to math away successes, because Shadowrun needed more math.

And all the extra work you have put in for players to do, to make their characters worse, still adds nothing fun to the game.

[luizborges]

Man if you fumble a simple dice roll this bad no wonder you don't like criticals and glitches...     

[ShadowcatX]

Nice job deflecting the issue. But given that you haven't addressed the basic question of why you want this I shouldn't be surprised. You have fun with it, I doubt your group will.

[Marcus]

Shadowcat's Question is reasonable, you haven't really said why you want this?
I played D6 Star wars for years, and I'm very familiar with the downside of that, and you version even lacks the small upside of it. There is a reason that system didn't go forward and the "wild die' was one of those reasons. I wouldn't want that rule active in SR game I was part of.

[Mirikon]

Indeed. That is far too complex, for far too little benefit. And this is coming from someone who plays HERO System.

[luizborges]

@Marcus and a few other, I already address why I think this is important "fix" the glitch system in this thread in a few places, added bold when needed:

Also, I don't mean to use glitches as crutch, I mean for that to happen as described in the book. As they are described in the examples they should happen with some frequency (you even have means to negate the glitch) because they are minor things, take a look in the book and check it for yourself.
(...)
So why have glitches in a system if they never happen? Either house rule them out (and stop worrying about 1s) or make they happen when they really should. I really think the devs "glitched" the glitch system. Maybe they didn't check the odds correctly or something like that, but as it is, the odds are a mess and I doubt anyone would want those odd curves on purpose...

3 glitches in 12 sessions, how is that relevant for your games? Do you really worry about glitches? Because with that record and considering 4-5 players you are looking a 1 glitch per player every 20-30 sessions... [ADDED INFO: this is in regard to the description of glitches mentioned above, that makes it looks like they are much more common]

I started this post because I researched a bit, run some probabilities and made a few tests using a dice roller to have a "practical" onlook on those odds. With a dice pool of 10 I got 2 regular glitches in 500 tests (no critical) and with a pool of 12 o got ONE glitch in 1000 tests (again no critical):

As the book described, it nowhere says glitches should be rare. It even has TWO mechanisms to deal with them, so it does expect them to happen. Also one of the ways to recover Edge is to endure a critical glitch. One edge is recovered every day with proper rest, so endure a critical glitch might be something stressful and hard, but as mentioned not something "rare as a blue moon".

This one is mostly opinion based, but I will talk about that after the quotes:

(...) your answer is one of the reasons I'm here discussing that. A glitch like a harmless blunder IS fun. It adds variety to perfect soldiers. You yourself said that you save the best ideas for players, but if the only ones who glitches are hapless incompetent NPCs, then what is the point? Here I'm talking about opinions I think the glitch idea is fantastic, but as it is, it is ineffective because the players rarely glitch...

Now back to my OPINION. In Shadowrun Anarchy (a "Fate" based version of Shadowrun), you don't have much extra in terms of success and failure, but you have the Glitch Die instead. The Glitch Die up the ante on test and can be added by anyone at any time just using a plot point (it is cheap to use it). With that die your test has 1/6 flat chance of glitching. Simple and direct. It also has a 1/6 chance of Exploiting if you succeed. Now lets think about this both ways:
- The highest probabilty of glitching in my system is with a pool of 0 dice, 1/6 chance of critical failure, it lowers quickly beyond that. My system make it looks like a glitch die is added every 4-5 tests, this would be much IF there was no means to counter it, but there are. You can pre and post edge with push the limit to avoid a glitch, you can use close call to negate a glitch, you can endure the glitch to get an edge point back. IF you glitch you have 3 ways of dealing with it.
- Now about the exploit side of things. This already exist in the system, even though it is not mentioned. Your net hits in many places determine how well you succeed. If you use Push the Limit or Second Chance you can bring that even higher. Now it just up to the GM to interpret that. If you made a test with threshold 1 and have 5 net hits, to me is not a simple success, it is a GREAT success. So the system is already in place, is up to the GM to use. And since my system focus on one small aspect of the system (the odds of glitching), there is no reason to mix stuff up and make a house rule that approaches both matters together.

This quote of mine is in context of my "fix" for the glitch system and its benefits, and why I would like that.

With those glitch mechanics critical glitches are still rare above 4-5 dice, but they do happen "once in a blue moon".
Regular glitches happen around 2-5% of the rolls above 4-5 dice.
Both (glitches and critical glitches) follow nice smooth curves (again to be added later), have no odd/even dice pool problems and no special cases whatsoever.

A few people didn't know or never noticed the special cases or problems, so I will explain those here:
- 0 Dicepool is a special case not covered by the rules. You can rule that you can't make a test like that, but in some situation like an opposed test there must be some chance to fuck it up by have 0 dicepool against someone, and not simply "i'm standing there doing nothing and staring at my opponent". With 1 dice pool the system has 1/6 chance of critical, and 0% with 0 dicepool? BAD!
- One thing that already poped up in this thread: a glitch happen with "more than half" ou "half or more"?? In 4th edition it was the second, now it is the first. Both of them have problems and are confusing...
- The RAW says that everywhere you have a division you must round the result up unless told otherwise. Also, the RAW of Glitches is "If more than half the dice you rolled show a one". Half the dice is a division, right? Most people interpret the glitch as split your 1s in two piles and if the 1s is the largest it is a glitch, BUT if you follow RAW strictly a glitch would happen if 1s are larger than POOL DIVIDED BY 2. So, A pool of 10 dice, glitchs with 6 or more, what about a pool of 11? Strict RAW would be 11/2=5.5 => 6. So you would need 7 or more dice to glitch.
- In RAW there is stuff like a dice pool of 3 is MUCH worst than a dicepool of 2 (about 3 times the odds of glitching), and the same happen with all even pools being worst than the even pool right below it.
- In RAW there is no probability curve, there is a jagged line. This IS terrible from any stand point.

NOW, I will addess ShadowcatX. I haven't seem a question in his last post. In the previous one he questioned (rethorically) that the system was more complicated and annoying, I explained him that it is not more complicated. The only thing "extra" it has is hit cancelation, that if someone isn't confortable, just remove it, the odds for criticals will drop a lot. My rule is one page long, it covers Description, Reasons for it, Mechanics (in just 4 lines), Clarifications (that is mostly retelling the same stuff in a way that it is clear), Comentaries (on ALL Edge and Quality effects that affects glitchs, mostly it is more clarifications that just says: "work as described" or add a little bit of extra information so that there is no doubt).
The rule could easily be:
"Roll 1 extra die, if THIS die is 1 and the total of 1 is more than hits, you glitched."
 or even
"If you have 1s equal or more to the numbers of hits, roll 1 extra die, if it is 1 you glitched."
Those would work just fine, without the hit cancelation, and will fix ALL the problems with RAW. The criticals would still be quite low and non-existent in higher dicepools, so I like the hit canceltion aspect of it.

About his example diceroll, why all the confusion? You throw 16 dice, and forget the glitch die, throw it aside and check for 1, any other result doesn't matter. If you read and understand the system, you will know if you have to throw the glitch die or not, because if you have less 1s than hits, the glitch die will at most give you one extra 1, so not enough for a glitch and not needed. About the "extra math", the SAME kind of thing is already used in Shadowrun:
- every two full boxes of excess Stun damage, carry over 1 box to the Physical damage track
- inflict 1 DV of Matrix damage to the target for every two full net hits
- every two full net hits counts as one hit on a Matrix Perception Test,
- every two hits adds one Initiative Die
- every two hits, increase the light penalty by one category

Those are just the ones I found searching "every two". It is EXACTLY the same as my system, "every two 1s over total hits cancel 1 hit" could also be read as "every two NET 1s cancel one hit". If doubt anyone playing Shadowrun is bothered or confused by this sort of "math".

[luizborges]

I played D6 Star wars for years, and I'm very familiar with the downside of that, and you version even lacks the small upside of it. There is a reason that system didn't go forward and the "wild die' was one of those reasons. I wouldn't want that rule active in SR game I was part of.

I looked into the D6 Star Wars system, and it is not in anyway like my system aside from the fact that it uses another die with your roll. The way it works is very much different, the odds of bad results are MUCH MUCH higher, the mechanics are horrible (depending on the result you add, subtract, add something and subtract another, etc).

[Marcus]

To be clear i'm not try to defend Shadowcat, they have their own issues, I was interested in why you wanted this and you did answer that question mostly I think far to concerned about small number cases.  I clearly missed something in your rules, as I understood it, if your glitch die goes one you glitch, as with Anarchy. I believe the industry term is Narrative based system. Why they felt the need to make Anarchy I honestly have no idea.  While there is nothing wrong with the Narrative concept, and it works fairly well for things like Fantasy Flight Star Wars, Shadowrun is a classic game with an accepted system. Why distract from what works? 

0 die pools don't exist, if you arrive at one in an extended test your done, extended test over. Otherwise it's minimum one. So I'm not really sure why that's even being discussed. If you roll a one on die pool of one you just critically glitched. There isn't some tricky rules lawyering argument that gets you out that, 1 is greater then 1/2 so you have more then 1/2 your pool showing as one and you have no successes therefor critical glitch. I don't think there is a GM in existence who was confused by this, and the GM who set this up is probably a jerk.

The Bell curve covering success vs failure is the probability curve that everyone looks at, and generally we are more concerned about the effects of limit on that curve then we are about the extremely unlikelyhood of a glitch. No one rolls a die pool of 1, or a pools of 3, why b/c players know better, if you take a look at how we recommend creating characters it will ensure such things don't happen, and if someone comes across such a situation then its ether edge time or just accept that you failed and try something else. That jagged curve is a non-existent one.

For the record I do think the rules say glitches are unlikely, it says so when it defines how glitches happen.