PDA

View Full Version : Thrasher Ability Breakdown: Probabilities



Veringatorix
02-28-2013, 08:12 AM
So I decided to break down the Thrasher ability and see what the chances of landing certain attacks are since there is a lot of discussion on his ability.

http://i47.tinypic.com/fmtm51.jpg

Chances:
4-7 str 4.17%
3 str/1-4 armor 4.17%
3 str 4.17%
3-6 str/1 armor 12.51%
2 str/2-5 armor 12.51%
2 str/1 armor 12.51%
2-5 str/2 armor 12.51%
1 str/3-6 armor 12.51%
1 str/2 armor 12.51%
1-4 str/3 armor 4.17%
4-7 armor 4.17%
3 armor 4.17%

Total: 100.08% .08 comes from truncation error

Base Damage:
3 str dmg chance is: 25%
2 str dmg chance is: 62.5%
1 str dmg chance is: 92%

The same calculations can be made for armor. I would point out that these base damage numbers simply look at str, there is obviously additional damage in the form of armor, but introducing a 2nd variable just complicates things.

By looking at the data I would say there is no chance for the Thrasher to be overpowered but on the flip side he has acceptable damage output even considering the chances involved. In short, he's fine as he is. Please feel free to discuss your opinions and point out any errors in my work. :)

djangoc
02-28-2013, 08:26 AM
Threshers are definitely useful in some builds.

Ziggy
02-28-2013, 09:27 AM
Think you missed one of the all armour sequences from the 100% calc (it should be 91.7%)

Veringatorix
02-28-2013, 09:46 AM
Yep, good catch, I had a typo and it translated down. thnx

franknarf
02-28-2013, 09:57 AM
Hrm, after Ziggy's complain I decided to burn five minutes in excel. There are 12 possibilities in total (four cases [0/3,1/2,2/1,3/0] for the early hits, three cases for the final hit), [so, what I said here was wrong! :)]

raven2134
02-28-2013, 10:05 AM
But what does it all meannnnnn? :p No, seriously, for the benefit of people colorblind and with basic arithmetic only.

Veringatorix
02-28-2013, 10:07 AM
Well, lets say .5^3 for the first 3 hits, then .33 for the last hit for all str. so 4 str hit is 1/24 = 4% so there are probabilities lower than 8%. :)

Ziggy
02-28-2013, 10:09 AM
But what does it all meannnnnn? :p No, seriously, for the benefit of people colorblind and with basic arithmetic only.

Veringatorix's summary is pretty much all you need to know. You'll almost certainly get at least 1 str damage, and are reasonably likely to get 2.

franknarf
02-28-2013, 10:09 AM
Ah, you're right. The combinatorial part of my brain is broken!

jofr
02-28-2013, 10:11 AM
But what does it all meannnnnn? :p No, seriously, for the benefit of people colorblind and with basic arithmetic only.
It basically means that the flail is a 90%-shot to make 1 str dmg. Everything else is just bonus.

Unless you want armor break, then it's a 90%-shot to make 1 arm dmg.

If he's maimed or you're out of ideas then it's a chance for him to do something usefull. :)

Veringatorix
02-28-2013, 10:14 AM
Ah, you're right. The combinatorial part of my brain is broken!

Haha, I like to do everything by hand so I can work through everything in my head. One of the reasons I dislike Excel for math stuff, everything is quite automated. :)

piotras
02-28-2013, 10:25 AM
Thanks for looking into it Veringatorix!

I wrote a few (in my view) constructive-criticism posts all over the place and I don't want to sound like a broken record, so I'll just summarise my own and a few other people's issues regarding the Trasher's Blood flail:

- it works and fills a special role, but many players are simply annoyed with how he executes his special ability (and I agree, if a player is not rewarded for pulling off a formation in which the Trasher is being flanked from all sides than something is very wrong)

- statistically speaking Bloody Flail is perfectly fine, but 'bloody flail's opposition fornt's (BFOF! :D) views are based on the few hits of bloody flail pare game (very often game-changing) rather than the hundreds of instances that would be required for the Bloody Flail to 'balance out' leaving one or the other player pissed off about the outcome of a game

- I would argue that we can still have him as a 'wildcard' unit build around randomness, but without the bittersweet mechanic we have now; i.e. what is important:

--- keep his identity; he was envisioned to be a wildcard, scary due to being unpredictable (to a certain extent)
--- reward the player for playing well,
--- in the spirit of Stoic's ideology of putting fun before gameplay and balance; do not make the player feel that he lost due to luck

My suggestion would be that bloody flail does 4 hits, with 25% base chance of attacking strength. For every adjacent ally, that chance would increase by further 25%. Most importantly: no extra damage from adjacent allies - in my opinion the single biggest issue which exponentially increased the damage output of almost pure-Trasher builds.

So when rushing solo he wouldn't be much of a threat (statistically-speaking!). When in a common scenario of a single ally next to him it would be the current 50/50 chance, while he would deal a guaranteed 4 damage when flanked by allies from 3 sides.

EDIT:

It would be great to capture the beauty of getting extra lucky armour-ignoring shots and promote tactical formations that increase your chances of success as it was in the early beta. Remember that in the eyes of player that misses the last hit it's a potential loss of 1-3 damage and the harder the player was working on strategic formation the more devastating it feels.

tnankie
02-28-2013, 03:37 PM
Piotras is right that BF annoys me as there is not statistical averaging on a per game basis.

There simply aren't enough rolls per game for any normalisation effects to occur. I've run coin toss simulators where 100s-1000s of flips are required to demonstrate the coin is fair (falls equally heads or tails).

However to address the OP, I haven't got the vibe that anyone thinks that the thrasher is OP. People are complaining that random chance is determining games. As demonstrated there is a wide spread of outcomes with the thrashers ability.

In the other thrasher thread (why do people keep starting new ones?) I suggested removing the hidden information by displaying the outcome of the next luck rolls to both players (next 5-7 rolls). I still believe this to be a good solution although implementation and UI could be very difficult.

However, this is about the thrasher.
How about 4 hits, every hit does 1 str or 1 arm (except the last hit gets current bonus for adjacency).
But rather than 50-50 every hit is an attempt to hit through armour, on fails do armour damage, if it succeeds do strength damage.

Cerzi
03-01-2013, 12:25 AM
Is it just me or is the Thrasher the most mis-read class name in gaming history? Seems the probability that they'll be called Thrasher, Thresher or Trasher is the same as the chance of them hitting strength, armor or missing that final blow.

raven2134
03-01-2013, 12:29 AM
Sky Seekers, Sterling Archers, Warbringers I've seen all sorts of stuff :p.

netnazgul
03-01-2013, 12:52 AM
Thrasher, Thresher or Trasher depends on how he is played mainly :)

Alex
03-01-2013, 04:29 AM
My favorite is that the game is frequently called "Fractions", even by press. I guess there is more math involved than most games.

Shiri
03-01-2013, 05:38 AM
Sky Seekers, Sterling Archers, Warbringers I've seen all sorts of stuff :p.

STERLING ARCHERS? How does that happen? I didn't think that many people were even familiar with the word sterling!

piotras
03-01-2013, 10:21 AM
...coming back to the topic:

Would anyone use a Thrasher as part of the core of their game-winning tactic?

I for one would never play a tactic build around WL's forge-ahead of SRM battling ram with Thrasher in mind, since it feels like wasting extra WP and turns when he misses (or attacks for armour). Heck, even my favourite 4-mixed-raider-formation-magics looses all the fun when a Thrasher misses a few times in a row.

If it was a question of always doing 4 damage, one or the other, it wouldn't be so bittersweet as it is now. Not being able to tell if I do 3-7 damage is just out of the question for me. The only adequate role I see for the Trasher is a village-idiot-type: soak-up damage, get some hits in and die - the sooner the quicker my turn advantage kicks in.

Ziggy
03-01-2013, 10:47 AM
Would anyone use a Thrasher as part of the core of their game-winning tactic?
I see a lot of WM/Trasher set ups (might be as I'm still messing about in rank 2-3), which are real a pain as leaving anything alive becomes a gamble.

netnazgul
03-01-2013, 11:17 AM
Thrasher is a great skirmisher and flanker, I prefer using him that way. It's a matter of taste and strategy to choose between BB and TH - they are both good at flanking, have a lot of movement and have good penetrational abilities (I love that 6 movement + 3WP hit!).
But you can't strategize solely on those units I think.

Ziggy
03-01-2013, 11:26 AM
But you can't strategize solely on those units I think.
In fairness, my favourite "strategy" so far is "run up with a Warhawk and hit Tempest".

jofr
03-01-2013, 11:55 AM
My favorite is that the game is frequently called "Fractions", even by press. I guess there is more math involved than most games.
Watch out for copyright infringement!
http://twinbeard.com/frog-fractions

raven2134
03-01-2013, 12:37 PM
I don't think it has to be a take it or leave it. True it's not a precision play unit - but I hardly ever played it like a precision play unit. I played it expecting 2 str damage - which the probability distribution shows to favor and for chip damage when maimed. I think people got used to playing the TH as if he was supposed to hit 3-4 str damage for 1 willpower consistently, but based on that conversion, it far surpassed many rank 1 abilities even saying that it's situational and requires formation.

piotras
03-01-2013, 01:22 PM
... but that was due to guaranteed extra damage on adjacent allies, which made him a definitive 3-4 armour-ignoring dmg per WP. Rather than removing that mechanic altogether it got even more complicated, with the outcomes very often being simply annoying rather than rewarding to the player.

Ehh, will give Thrashers one more try but the more I play the more I hate this unit (and I don't use that word often). What suck even more is that it was my favourite class and at least a single TH was a must in my (mostly raider-heavy) builds, now I just can't stand the guy.

raven2134
03-01-2013, 01:33 PM
Haha...if Janus had SA's I guess it was TH for you :p.

piotras
03-01-2013, 02:02 PM
Haha...if Janus had SA's I guess it was TH for you :p.

kinda :) I never played the famous 2x WM 3-4 TH build however :) But I loved to have a good mix of raiders, and 1 or 2 TH were a must. I think I enjoy the formations etc. much more than range play with archers. I also loved the 2xRM, 2xTH, 2xSTR build.

I guess that's why I get so annoyed with thrashers being so unreliable. What these statistics don't get into account are variables like: how often do you miss 1 or 3 damage? Missing 1 damage when you're solo is less annoying than trying to synergise your other raiders tasks like armour break or archer harassment for the past few rounds to work in concert with the TH just to see him miss on the last 3-point shot OR do 3 break on the last hit to a dude who is already stripped off armour but gets it from the raiders shield wall passive... I know what I'll do, I'll have a piece of paper of my own and see how many times my blood boils when I see the outcome of my thrasher's hits :p

Honestly, this is so annoying I will probably stop playing raider-based builds because TH was a like the centrepiece of it.

RobertTheScott
03-01-2013, 03:16 PM
Watch out for copyright infringement!
http://twinbeard.com/frog-fractions

Frog Fractions is EPIC!