Includes a standard deviation score in PBEM

[Paul806]

Is it possible to have a tally of deviation in combat result in PBEM ?

Simple version :
- each time a side makes more losses than expected, and/or takes less losses than expected add +1 (to a maximum of +2 per combat)
- each time a side takes more losses than expected, and/or makes less losses than expected add -1 (to a maximum of -2 per combat)

Advanced version :
- multiply the standard score by MPP value of the loss

It's a simple tool to gauge a player's honesty : if someone reloads the turn multiple times until he gets a good result, the score will go up.

Also : is it possible to hide the changes in US - USSR mobilization rolls until next player's turn ? Again to avoid a reload of the turn if you have a bad result for your side.

I don't know if these changes are easy to implement, but I think they can go a long way to insure honest play.

[OldCrowBalthazor]

@Paul806
The PBEM server is supposed to flag repeated 'reloads' of a turn, though I know the 'flag' (an email warning initially) must come up only after quite a few reload attempts.

I don't cheat and personally hate cheaters, but know that the warning mail from Slitherine doesn't happen if a few reloads are done in a single turn because of experiences I personally had of repeated and temporary internet glitches because of storms.
(Note I and quite a few other folks I play with will communicate to their opponent if they had a turn interrupted because of having to reload a dropped turn)

One result of having to reload an interrupted turn is that new Intel rolls are made, and different reveals can happen...a real bummer if your a sportsman.

Just curious what prompted your work around? It sounds like you suspect you have had been cheated, and more than once.

It's an interesting idea that you posted here but there are legitimate reasons why folks have to reload a turn...at least in my experience.

[Ktonos]

I got only once such a message before, after a sole reload of a turn (initial play crashed on Pbem error). The thing is, does the opponent get it too? ie was my opponent notified that I reloaded the turn? Because if not, then its a moot mechanic.

I also experienced strings of "perfect results" from opponents, tanks attacking entrenched armies never getting a single casualty until they become elite,... or my fleet ambushes/maneuvers always getting discovered by others, but it is a big step to nothingness to suggest to someone you suspect foul play. Another moot action: a cheater will always deny and an innocent will be offended.

Now to have a sum up of the expected deviation I don't thing will do any good. I find myself doing above the predictor when the war goes well, and below when it goes bad. There may be some NM algorithm assosiated with it.

Maybe a workaround would be to mark the game on the lobby with italics when it has been loaded rather than saved. Or even better have a number next to it, depicting the number of times it was loaded by the opponent.

[Bavre]

It is difficult to tell with combat results, as you can't see the enemies odds, however there are a few easy to spot telltale signs (because they have fixed odds). A few examples:

GB stratbomber evades damage twice a round for a dozend+ rounds straight in a row.

Port attacks by capital ships do collateral to the ship in port >90% of the time.

During the battle of France, GB carriers do damage to land units with every single attack over several rounds. Note that this does not apply to Jap carriers! GB ones have naval weapons 1 and 0 XP, so their chance to damage land units is initially never more than 33%, no matter how weak the target. Japs have the same tech, but two of them start at 1 XP and therefore can get reliable 0:1 odds vs weak land units.



Actually I really like this idea with the deviation. My statistics lectures were admittedly ages ago, but afaik getting so lucky that it seems you're cheating is nearly impossible. Some of the most extreme "lucky players" routinely have rolls the equivalent of multiple lottery wins each match.


@ Ktonos:
Reloading the odd turn does have no effect at all. Living in a very remote rural region, I've had my share of disconnects just like OCB. Even 5 or 6 reloads a turn due to a persisten (and reported) bug caused no reaction.
The only time I ever got this warning email was during the server problems somewhere late May/early June and there I must have downloaded one and the same turn at least a dozend times over the space of some days, as it just kept crashing. This warning also was just a very polite email, inquiring whether I am having internet trouble.

[ThunderLizard11]

LIke it - shouldn't be that hard to add at it can display as a running counter for each side.

[Paul806]

One result of having to reload an interrupted turn is that new Intel rolls are made, and different reveals can happen...a real bummer if your a sportsman.

This also can be easily solved : make the intel rolls at the end of the other player turn will avoid any variation. That's the solution I proposed for entry rolls.

One of my best PBEM experience was with the old EastFront game by Columbia games. Why ? Because all rolls were made at your turn, but the results were visible only after you sent the turn. Obviously this can work only when the combat phase happens at the end of the turn, but it was impossible to cheat at this game.

Just curious what prompted your work around? It sounds like you suspect you have had been cheated, and more than once.

Bavre and Ktonos answers are giving good exemples of some extreme results. Let's just say I had my share

Now to have a sum up of the expected deviation I don't thing will do any good. I find myself doing above the predictor when the war goes well, and below when it goes bad. There may be some NM algorithm assosiated with it.

What I propose to mesure is only the random part of the result. According to the manual :

In addition to each of the factors described above, there is a possible +/-1
point variation in each combat that provides some randomness to the
combat calculations

There is a snowballing effect in the battle mechanism which I rather like (a nice way to model momentum and initiative), but it is also an incentive to dishonest play.

tanks attacking entrenched armies never getting a single casualty until they become elite

Yeah, it is really a problem and for me the biggest telltale of foul play. An honest player will almost never attempt to attack a well entrenched enemy. I see two fixes possible :
- calculate also the deviation for the fortification rolls
- add half the value of the fortification if you miss the first roll. So the probability of ignoring a 6 level of fortification two times in a row drops to 4% instead of 16%

Another moot action: a cheater will always deny and an innocent will be offended.

Now you'll have a mesurable gauge. A player who has always positive scores in his games is either very lucky (and should play the lottery) or dishonest. I never accused an opponent of anything : as you point, this is moot.

Actually I really like this idea with the deviation.

Thank you. And thank you all for your input.

[ThunderLizard11]

OK Devs- what do you think? Can this be added?

[OldCrowBalthazor]

OK Devs- what do you think? Can this be added?

Yeah I'm wondering too. Seems cheating (with MP) is a bigger problem than I thought. Several people told me that basically the described shenanigans with the log-out deal has been a thing for a long time with the PBEM server. And not just with SC games.
It's possible that this is something that Slitherine needs to hear, if they haven't already.

[ElvisJJonesRambo]

Really want to bring this up again?
There's many threads throughout.
We know how this rolls, pun intended.
It's like discussing steroids in sports.

Recording odds is interesting, but waste of time.
"Everyone is an expert" on statistics, just goto a casino.
Bunch of losers, sitting around a Craps (Dice) table, Roulette Wheel.
"This game is rigged, how does it keep hitting 7?

Even if you get "good" or "bad" luck, it's like a football game, "the one important roll matters most".
(Recording Garrison vs Garrison stats, who cares.)

[ThunderLizard11]

Really want to bring this up again?
We know how this rolls, pun intended.
It's like discussing steroids in sports.

Sometimes it takes a while for the rest of the group to catch-up:

https://www.matrixgames.com/forums/view ... g#p4915141

[ElvisJJonesRambo]

The real problem with Matrix, Slitherine, or whatever multiple names this forum has is something else.
The problem is the Forum is boring, 95% of the people are boring.
No sense of humor, constantly offended.

Have had several douchebag(s) "get offended" with messaging within the game.
Who walks around life, getting offended every 30 seconds.
All over a "Nothing Burger".
There's no vibe. Censorship sux.
Gonna (not) party like it's 1984.

"Colonel, did you reload the turn?"

[ElvisJJonesRambo]

@Paul806 --- With 8 posts, and never seeing you host a game, or even breathe, who have you played? How do you get PBEM games without posting? Or you secret PM guy? Generation "Just PM me".

I did PM you. Simple, wanna play a game, got an open spot?

[Hubert Cater]

OK Devs- what do you think? Can this be added?

Thanks for the posts everyone and while it is unfortunate that players will find ways to cheat, I'm not sure if there really are any easy fixes here outside of what is already in place.

Reloads counter - technically this could be displayed, but my understanding is that this is not officially desirable as there could be legitimate reasons why a turn was reloaded, and perhaps even several times during play. Abusers should be warned and sometimes kicked under the right circumstances (which I have seen from my end), but the general desire is to not end up in a situation of possible public shaming when the reloads counter says one thing, and players potentially disagree on the legitimacy of the reloads or not.

Randomness - Yes this could be locked in at the end of a turn, for the next players turn, but I could see the potential for abuse here as well, e.g. as a simple example, if you play your turn and got a +1 for your first three or four attacks, you could reload your turn and ensure you use the +1s for critical attacks where those +1s make the most difference. I can just picture multiple ways to game the system when you know the order of the +/- 1s for attack results.

Spying Reveal - Yes, that could be amended by locking in at the end of the previous turn for the next turn, and off hand I don't see a way to game that easily.

I know there is more, but those are a few off the top of my head where I can see some potential changes that might help, but I doubt there will ever be wholesale changes that completely secure the game from abuse when players reload their turns, unfortunately.

[Ktonos]

I always assumed that everyone's play routine was load a new turn and finish it in one sitting. At least thats my routine. Almost all the reloads I had were due to a Pbem error at the end turn. In all these years of playing SC games, there are like 2-3 maybe 4 times I got to save mid turn because of an emergency.

I also usually inform the opponent - if we are already exchanging pms that is - that I had a Pbem crash (and I get notified also).

Load counters would only mean smthing if accompanied with weird combat results. Though I get that this might end up stirring players to paranoia if they are losing a game fair and square, but they see a big load counter on the opponent.

People could also sit down and agree beforehand if the "reload counter is on", especially if the bulk of the people here are indeed a one sitting players.

But yeah, in the end, if someone is gonna cheat, he is gonna just do it. But there could be some rules enforced for the Pbem games.


Edit: Wouldn't you agree, that if the Reloads are many, but the Saves few, there's something fishy?

[Feinder]

I had no idea any of this was even possible.
The only PBEM opponent I've ever played is my brother, Ngineer. And he's kicked my butt two-for-two.
I do "save early an often" to he PBEM server.
I will often NOT be able to complete a turn in one sitting (that's what the Save and Quit buttons are for, correct?).

I guess my response is, "Hm. Interesting. I never knew this reload issue, was an issue."
I expect Ngineer is in the same situation I'm in (need to save-exit-return at a later time).
There's no nefarious intent tho. Just the need to attend to the distractions of Kids, Bio-Breaks, Wives, Eating, and Jobs.

I do take my game(s) with Ngineer seriously. But it shocks me to hear/read the lengths that (other) people will go to game the system. I mean, I know it happens all the time in real-life. But this is just a game, and it's not like there's any money riding on it... Oh well. Give folks the anonymity of the internet, and they'll try to get away with just about anything.

[ThunderLizard11]

Thanks for the posts everyone and while it is unfortunate that players will find ways to cheat, I'm not sure if there really are any easy fixes here outside of what is already in place.

Here's my strawman:

* Move calcs for all RNG events like diplo, spotting, tech advances, etc. to the end of the prior players turn but report results at beginning of next players turn - that way they are locked.
* Add a reload counter - to distinguish those who save/exit/save - a reload is only the subsequent save of the same load file. For example, you load turn, make a few moves, save/exit, come back later and do the same, etc. - reload counter is still 1. You load game, make some turns, your PC crashes, reload counter = 2.
* Add a RNG metric. Average result = 0. Better than average = 1, 2, 3 depending on how much better than odds player performed. Overall a long game you might see a slight up and down but average should meander around zero. Other option is to tone-down/turn-offf RNG altogether (maybe option?).

The other thing I'll mention, at times it seems some players are playing with fog-of-war off. For example, Brits happen to land a AVL in Bremen while Germanys are busy down south with French.

[Ktonos]

* Add a reload counter - to distinguish those who save/exit/save - a reload is only the subsequent save of the same load file. For example, you load turn, make a few moves, save/exit, come back later and do the same, etc. - reload counter is still 1. You load game, make some turns, your PC crashes, reload counter = 2.

That's gold I think. Yes, why reload the same save, bar for pbem errors?

Also, I don't think that warning e-mails would stop a cheater. I think, a really strong deterant, if not the sole one, would be for the cheater to know that you know.

[ElvisJJonesRambo]

Come on get real.
I lost my first 30+ games in ELO.
Then I figured out the Scooby-Doo mystery.
People reload.
I'm guilty.
Then I started winning a few.
Terif cheated for 2 years, in SC-1, until we figured out how he did it.
Even when playing live TCP/IP.

Talking about this, is like the steroids in sports. Women's Tennis, Soccer, Football, Baseball, UFC, Lance Armstrong, etc.
Every sport is cheating.

For all have reloaded and come short of the glory of gaming perfection.

The clowns that say, "Oh, I don't reload, it's only when I'm playing in the backfield of my farm, using Skynet Internet. Also, it snows alot." (dude is in Arizona, Summertime).

Every see someone do something dumb with Paratroops?
HQ's sail all over the oceans, never getting hit.
Double Diplo hits in 1941 on DEI. (big hits, when I got maximum as deterrent)

Franky, I don't care. I'll post my loss. I just returned to computer games to escape the Covid scam.

I'm gonna be dead in few years (or less), so who cares.
I just want some smack and laughs. And some gaming.

[ThunderLizard11]

Come on get real.
I lost my first 30+ games in ELO.
Then I figured out the Scooby-Doo mystery.
People reload.
I'm guilty.
Then I started winning a few.
Terif cheated for 2 years, in SC-1, until we figured out how he did it.
Even when playing live TCP/IP.

Talking about this, is like the steroids in sports. Women's Tennis, Soccer, Football, Baseball, UFC, Lance Armstrong, etc.
Every sport is cheating....

Well Lance finally did get his comeuppance. And the Tour de France is much better off for it IMHO. Same deal with the East Germans of old.

[Bavre]

I think one should keep in mind the magnitude of reloads here. As said above I live in a very remote area and rarely finish a game without 2-3 disconnects/PBEM server errors. But 2-3 reloads, or let's be generous and say half a dozend, even when specifically used for cheating would absolutely not be noticeable over the course of an entire game. Some of those "miracle turns" alone must have taken many dozend attempts. A counter would of course not stop a clever and restrained cheater, but it would curb the worst excesses.
It might even eliminate some false positives, e.g. when I started with SCWW1 I was 100% convinced that something was fishy with some guys naval game until I figured out how scouting with silent subs worked. Some of my previous opponents probably had a similar feeling about my gameplay during WaW Barbarossa back in the day before spyfighters were nerfed, as they were simply not aware that I had vision range 5 or 6. A counter would have allayed suspicions in both cases.