Board "shake up" abilities (100% more Sniper Rifle)
A little while ago, whilst discussing C.Mags I was a bit stuck on how valuable board shake up abilities were (looking at his red) so I tested it out some (averaged almost exactly 2/3 of a match per use) and it got me to thinking about other abilities, how much they actually achieve and how valuable the shake up part of an ability really is.
So to break up the monotony of the TaT PvE i've been (over)using L.Thor, Hulk and Ragnarok to see how their board shake up abilities play out.
At time of testing all abilties had 5 covers apart from Mjolnir's Might at 3 covers but it only scales damage after than point anyway. (Thanks for pointing out I needed to clarify this Mehhh)
Hulk Anger
116 total uses of Anger some of which were consecutive uses after multiple hits (not many though). This resulted in 51 instances of no matches at all and 65 of 1 or more matches. Lowest cascade was 1 (obviously
) and highest was 6. 1 was the most common number and frequency reduced as cascade size increased.
The 65 uses providing 1 or more matches produced a total of 101 matches, mostly 3 but as high as 8. Total matched tiles was 327. This means each use of Anger provides, on average, 2.8 matched tiles.
Hulk Thunderous Clap
184 total uses of thunderous clap as close to 10 green AP as possible each time. Of those 65 produced no matches and 119 produced 1 or more matches. Lowest cascade was 1, highest was 8. 2 was actually the most common number of matches, followed by 1 then less frequent as you increase from 3 onwards.
The 119 uses providing 1 or more matches produced a total of 263 matches, mostly 3 but as high as 6. Total matched tiles was 825. this means Thunderous Clap provides, on average, 4.5 matched tiles.
Thor Mjolnir's Might
Pretty unexciting (as expected). 59 total uses of which a painful 49 produced no matches. Of the 10 that provided matches only 1 instance was greater than 1 (2!
) for a total of 33 matched tiles which equates to 0.5 matched tiles per use.
Thor Thunder Strike
35 total uses of Thunder Strike. Of those 6 produced no matches and 29 produced 1 or more matches. Lowest cascade was 1, highest was 5. 1 was the most common number of matches, then less frequent as you increase from 2 onwards.
The 29 uses providing 1 or more matches produced a total of 55 matches, mostly 3 but as high as 8. Total matched tiles was 183. this means Thunder Strike provides, on average, 5.2 matched tiles.
Ragnarok Godlike Power
64 total uses of Godlike Power. Of those 30 produced no matches and 34 produced 1 or more matches. Lowest cascade was 1, highest was 4. 1 was the most common number of matches, then less frequent as you increase from 2 onwards.
The 34 uses providing 1 or more matches produced a total of 54 matches, mostly 3 but as high as 7. Total matched tiles was 178. this means Godlike Power provides, on average, 2.8 matched tiles.
Ragnarok Thunderclap
70 total uses of Thunderclap. Of those 36 produced no matches and 34 produced 1 or more matches. Lowest cascade was 1, highest was 5. 1 was the most common number of matches, then less frequent as you increase from 2 onwards.
The 34 uses providing 1 or more matches produced a total of 40 matches, mostly 3 but as high as 5. Total matched tiles was 131. this means Thunderclap provides, on average, 1.9 matched tiles.
Loki Illusions
60 total uses of Illusions. Of those 27 produced no matches and 33 produced 1 or more matches. Lowest cascade was 1, highest was 6. 2 was the most common number of matches, 1 next then less frequent from 3 onwards.
The 33 uses providing 1 or more matches produced a total of 81 matches, mostly 3 but as high as 5. Total matched tiles was 255. this means Illusions provides, on average, 4.25 matched tiles.
GSBW Sniper Rifle
54 total uses of Sniper Rifle. Of those 17 produced no matches and 37 produced 1 or more matches. Lowest cascade was 1, highest was 11! 2 was the most common number of matches, 1 next then less frequent from 3 onwards.
The 37 uses providing 1 or more matches produced a total of 98 matches, mostly 3 but as high as 6. Total matched tiles was 310. this means Sniper Rifle provides, on average, 8.4 matched tiles.
Summary
Mostly some food for though on the value of these powers. Thunder Strike being amazing was instinctively obvious, same with Thunderous Clap but I was surprised how much work Godlike Power got done. Always though it was pretty bad but it was giving almost half it's cost back in random AP which is not to be sniffed at. Thunderclap returning just under 1/3 of its cost was also a pleasant surprise.
The less said about Mjolnir's might the better. Probably should perform a bit better but because I prioritize yellow so highly with L.Thor there's probably not all that much there for the tiles to match with. Similar situation with Anger to some degree I imagine.
Illusions is kind of meh. It's an under-represented colour which is a plus and you can hope for a good match after use but don't get your hopes up for an influx of AP after a use. Almost half the time it matched all of nothing for me.... I will say that on every occasion I had a crit tile on the board Ilusions at least had the decency to match it for me
.
Sniper rifle, on the other hand, is pretty amazing for board shakeup. You can target tiles to be destroyed (I can often pick three or more goon CD tiles off in PvE for example) and it supplies a healthy return of AP plus match damage alongside it's high AoE damage output (at 5 covers). Admittedly a fight is often pretty much over when sniper rifle goes off but it's still a potentially valuable bonus to the ability and in her popular Hulk combo you can get the rather daunting 19 tile cost fairly quickly.
So to break up the monotony of the TaT PvE i've been (over)using L.Thor, Hulk and Ragnarok to see how their board shake up abilities play out.
At time of testing all abilties had 5 covers apart from Mjolnir's Might at 3 covers but it only scales damage after than point anyway. (Thanks for pointing out I needed to clarify this Mehhh)
Hulk Anger
116 total uses of Anger some of which were consecutive uses after multiple hits (not many though). This resulted in 51 instances of no matches at all and 65 of 1 or more matches. Lowest cascade was 1 (obviously
) and highest was 6. 1 was the most common number and frequency reduced as cascade size increased.The 65 uses providing 1 or more matches produced a total of 101 matches, mostly 3 but as high as 8. Total matched tiles was 327. This means each use of Anger provides, on average, 2.8 matched tiles.
Hulk Thunderous Clap
184 total uses of thunderous clap as close to 10 green AP as possible each time. Of those 65 produced no matches and 119 produced 1 or more matches. Lowest cascade was 1, highest was 8. 2 was actually the most common number of matches, followed by 1 then less frequent as you increase from 3 onwards.
The 119 uses providing 1 or more matches produced a total of 263 matches, mostly 3 but as high as 6. Total matched tiles was 825. this means Thunderous Clap provides, on average, 4.5 matched tiles.
Thor Mjolnir's Might
Pretty unexciting (as expected). 59 total uses of which a painful 49 produced no matches. Of the 10 that provided matches only 1 instance was greater than 1 (2!
) for a total of 33 matched tiles which equates to 0.5 matched tiles per use.Thor Thunder Strike
35 total uses of Thunder Strike. Of those 6 produced no matches and 29 produced 1 or more matches. Lowest cascade was 1, highest was 5. 1 was the most common number of matches, then less frequent as you increase from 2 onwards.
The 29 uses providing 1 or more matches produced a total of 55 matches, mostly 3 but as high as 8. Total matched tiles was 183. this means Thunder Strike provides, on average, 5.2 matched tiles.
Ragnarok Godlike Power
64 total uses of Godlike Power. Of those 30 produced no matches and 34 produced 1 or more matches. Lowest cascade was 1, highest was 4. 1 was the most common number of matches, then less frequent as you increase from 2 onwards.
The 34 uses providing 1 or more matches produced a total of 54 matches, mostly 3 but as high as 7. Total matched tiles was 178. this means Godlike Power provides, on average, 2.8 matched tiles.
Ragnarok Thunderclap
70 total uses of Thunderclap. Of those 36 produced no matches and 34 produced 1 or more matches. Lowest cascade was 1, highest was 5. 1 was the most common number of matches, then less frequent as you increase from 2 onwards.
The 34 uses providing 1 or more matches produced a total of 40 matches, mostly 3 but as high as 5. Total matched tiles was 131. this means Thunderclap provides, on average, 1.9 matched tiles.
Loki Illusions
60 total uses of Illusions. Of those 27 produced no matches and 33 produced 1 or more matches. Lowest cascade was 1, highest was 6. 2 was the most common number of matches, 1 next then less frequent from 3 onwards.
The 33 uses providing 1 or more matches produced a total of 81 matches, mostly 3 but as high as 5. Total matched tiles was 255. this means Illusions provides, on average, 4.25 matched tiles.
GSBW Sniper Rifle
54 total uses of Sniper Rifle. Of those 17 produced no matches and 37 produced 1 or more matches. Lowest cascade was 1, highest was 11! 2 was the most common number of matches, 1 next then less frequent from 3 onwards.
The 37 uses providing 1 or more matches produced a total of 98 matches, mostly 3 but as high as 6. Total matched tiles was 310. this means Sniper Rifle provides, on average, 8.4 matched tiles.
Summary
Mostly some food for though on the value of these powers. Thunder Strike being amazing was instinctively obvious, same with Thunderous Clap but I was surprised how much work Godlike Power got done. Always though it was pretty bad but it was giving almost half it's cost back in random AP which is not to be sniffed at. Thunderclap returning just under 1/3 of its cost was also a pleasant surprise.
The less said about Mjolnir's might the better
. Probably should perform a bit better but because I prioritize yellow so highly with L.Thor there's probably not all that much there for the tiles to match with. Similar situation with Anger to some degree I imagine.Illusions is kind of meh. It's an under-represented colour which is a plus and you can hope for a good match after use but don't get your hopes up for an influx of AP after a use. Almost half the time it matched all of nothing for me.... I will say that on every occasion I had a crit tile on the board Ilusions at least had the decency to match it for me
.Sniper rifle, on the other hand, is pretty amazing for board shakeup. You can target tiles to be destroyed (I can often pick three or more goon CD tiles off in PvE for example) and it supplies a healthy return of AP plus match damage alongside it's high AoE damage output (at 5 covers). Admittedly a fight is often pretty much over when sniper rifle goes off but it's still a potentially valuable bonus to the ability and in her popular Hulk combo you can get the rather daunting 19 tile cost fairly quickly.
0
Comments
-
Thanks for this, very informative, for a while now I've been basing my team on abilities that create cascades, (Sentry/Hulk/The Punisher, C Magneto, The Hood), cascades help tremendously when you want to get rid of special tiles but can't seem to find some nearby matches. It's nice to see some numbers, I already knew how good The Hulk was from using him for months now, but wow L Thor numbers are insane, I need some more ISO to level him up0
-
Without mathing it up, Thor is definitely only second in cascading to MStorm so it'll be interesting to see how good lazy storm is at doing it too if her mistress cost doesn't decrease.0
-
Fantastic work!
Be interesting to see the numbers for C.Mags' red and Storm's green (either Storm's green).
I've always suspected that if he had a 3rd ability, or even if his abilities weren't red/green and thus overlapping with everyone, even the nerfed Rags is a pretty good character. It's just the difference between stupidly-overpowered Rags and current Rags is so high.0 -
Isn't GSBW considered a better cascading character? All 3 of her abilities do shake-up.HairyDave wrote:Without mathing it up, Thor is definitely only second in cascading to MStorm so it'll be interesting to see how good lazy storm is at doing it too if her mistress cost doesn't decrease.
Also MMN, but yeah, 2/3 CD abilities.0 -
Zhirrzh wrote:Fantastic work!
Be interesting to see the numbers for C.Mags' red and Storm's green (either Storm's green).
I've always suspected that if he had a 3rd ability, or even if his abilities weren't red/green and thus overlapping with everyone, even the nerfed Rags is a pretty good character. It's just the difference between stupidly-overpowered Rags and current Rags is so high.
From the Classic Mags thread...bonfire01 wrote:-snip-
Purple first then. A few times I waited a turn or 2 to avoid making almost no matches (I still wanted to win my games ) but over the course of 109 uses of MT I made a total of 784 tile swaps for an average of 7.2 swaps per usage. Lowest was 5, highest was 12. Most of the matches cluster between 6 and 9 with 7 and 8 having the highest numbers of times appearing.
Red next. 282 uses of C.Mags red in total of which 193 produced no matches at all after the tiles were destroyed. Of the remaining 89 uses I made a total of 189 matches, lowest being 1 match (unsurprisingly) and highest being 7. As expected 1 match was most common then 2 matches etc etc.
So overall for magnetic projectiles 282 uses produced 189 3 or 4 matches (mostly 3) for a total of 589 AP generated which is pretty close to 2 AP per use on average (2.08). So at 5 red you could almost say the ability is free... kind of. If the cost goes up to 5 as most ppl expect (but who knows what D3 will do) and the number of tiles destroyed stays the same then you could almost say it's real cost will be 3AP (although some of those matches will be environmental tiles or colours you can't use). That makes me think that a bonus damage value of something like 500-600 would be about right to give a reasonable damage per AP for a 141 but lower than pure damage abilities because you are in theory getting 2 random AP back and might destroy the odd special tile (low odds but at least they are there). OFC you might also destroy your own special tiles....
-snip-0 -
That's true but they're also a lot more expensive so I'd rate Storm as the better board smasher. That said, Widow is definitely better at targeted demolition.locked wrote:
Isn't GSBW considered a better cascading character? All 3 of her abilities do shake-up.HairyDave wrote:Without mathing it up, Thor is definitely only second in cascading to MStorm so it'll be interesting to see how good lazy storm is at doing it too if her mistress cost doesn't decrease.0 -
I mean, better than Thor. Storm is no doubt the best cascading character besides Magneto.0
-
Widow is probably better than Thor but he seems to get real lucky with that Thunder Strike.0
-
Love the analysis, well done sir!0
-
This is absolutely outstanding stuff. Thanks for sharing!
(As an aside, it makes me sad that complaint threads go green and this thread has just 4 upvotes. Seriously?)
I did some research on cascades myself a while back, limiting sample size for each to 20. My results:
14 random tiles (L5 Hulk at 12-13 AP, L5 CStorm at 14, mStorm): 1.45 matches, 3 criticals
13 random tiles (L5 Hulk at 10-11 AP): 1.8 matches (4.8 tiles, slightly higher than yours), 5 criticals.
5 random tiles (Magneto red): 0.6 matches (1.8 tiles, slightly less than yours), 1 critical.
Godlike Power: 0.95 matches (2.85, same as yours), 4 criticals.
Sniper Rifle: 2.9 matches (8.7 tiles), 7 criticals.
Judgment: 0.55 matches, 0 criticals.
Obviously we need more data, but some early analysis:
- Adding specific colored tiles to the board has a linear regression of -1.27 * 0.74 x. In English, this means that adding 3 or fewer tiles to the board is basically worthless in terms of generating cascades; beyond 3, you expect 3/4 AP for each tile you add.
- Shattering random tiles has a linear regression of -1.64 + 0.46x -- 4 or fewer is worthless, add about half an AP generated per tile shattered.
- Shattering chunks of contiguous tiles generates way fewer cascades than random tiles. It's likely that the cascade effect of Judgment is overrated.
- Sniper Rifle, on the other hand, looks underrated with regard to cascades, as it generated nearly 2 more AP than expected for random tiles.0 -
Nice work, bonfire! How are you doing all these statistics and *still* beating my score in TaT? 0
-
How does Loki, having pure 'shake' move with no damage, compare to those?0
-
Nivrax wrote:How does Loki, having pure 'shake' move with no damage, compare to those?
I've not done enough uses of illusions yet to be sure (don't use him all that much so takes time to collect numbers) but so far it looks like he'll average somewhere in the region of 5AP gained per use.0 -
mischiefmaker's character rankings from March say that it averages 4 tiles matched:
viewtopic.php?f=14&t=5093#p851760 -
The 'add tile' abilities is context sensitive. You can basically say any time you have more than 15 of the same color on the board then there's almost no possible way you can place them without having some kind of cascade or an available match 5, so any tile added beyond 15 is almost certain to refund that AP. That is, if there are 14 green tiles on the board and you did Thunder Strike, you expect to get at least 8 of those back just because there's almost no way those tiles could've fit on the board without cascading. Now the choice of 15 is fairly arbitrary but 15 is close to 1/4 of the board which is why I picked that, and I can't think of the last time I saw 15 tiles of the same color on the board that I wasn't able to turn it into at least a match 4 immediately (and usually much better). Now if you're going to add tile and have a number of under 15 then it's trickier, and it also depends on how the tiles are clustered. If there are already nearby matches then it's going to be more likely to have cascade versus tiles evenly spread out in the board.
For board shakeup ability there's an upper limit of how many cascade you get somewhere around the teens. I remember running a 30 green AP Lightning Storm with Classic Storm and was surprised to find that I didn't get much cascade, but then I realize that a 30 random tile shatter is awfully close to just the same as getting a brand new board, which is very close to Illusions. What's not clear to me is does the number of cascade generated by a random tile shatter ability plateaus at some point, or does it actually go down at some point. After all, Illusions is roughly the same as shattering the entire board, but Illusions sure doesn't seem to be as good as the old 16 tile Lightning Strike. Well, we can be pretty sure Illusion sure isn't better than a 16 tile Lightning Strike, but whether it's worse is difficult to analyze without a ton of data.0 -
Yes there would be a natural plateau to the effect of increasing numbers of shattered tiles. The mathematics behind it is way to complicated for me
but a simple explanation would be this:
If I shatter the tile at the very bottom of any given column then 7 tiles will move down 1 spot and 1 new tile drops onto the board. Any one of those 8 tiles have the potential to make a match thanks to the movement.
If I instead shatter the bottom tile and the one above it then 6 tiles will move down 2 spaces and 2 new tiles drop on. This still gives me 8 new tile placements which have the potential to make a match and the fact 2 new tiles are dropping on instead of one only very slightly improves my chances.
Any configuration of 2 tiles shattering in the same column is, to differing degrees, less effective than a single shattering at the lowest point of the 2 plus a random shattering in a different column.
So basically, as you shatter more and more tiles you'll increase the likelihood that multiple tiles are shattering in the same column and thereby be gaining less per additional shattered tile BUT there is a lot of room for extra tiles to improve your matching rate because you increase the odds of shattering lower placed tiles.
As to when you start getting an appreciable diminish in returns .... my maths skills aren't up to that one.0 -
Given that we know there are only seven types of tiles (wrt to matching), and the educated guess that each of them is equally likely, it ought to be not too hard to write a simulator offline to test these theories, thousands of times if needed.
(Something for me to thinking about while jogging today.)0 -
One thing I'm missing in the OP, bonfire1: how many tiles were replaced / dropped in? It's easy for Godlike Power (16 in the middle), but what level is your Hulk's anger, thunderous clap, etc.? It matters, and I didn't see mention of it...0
-
Mehhh wrote:One thing I'm missing in the OP, bonfire1: how many tiles were replaced / dropped in? It's easy for Godlike Power (16 in the middle), but what level is your Hulk's anger, thunderous clap, etc.? It matters, and I didn't see mention of it...
All abilties were at 5 at the time of testing (although I specced Hulk to 5 red since and am now in the process of putting him back to 5 green 8) ) with the exception of Mjolnirs might which is only increasing damage IIRC not tiles.
Some of Hulk's claps will have got an extra tile or 2 due to green AP but I was clapping as soon as I could after reaching 10 AP for testing. So only green cascades really messed it up.0
Categories
- All Categories
- 44.8K Marvel Puzzle Quest
- 1.5K MPQ News and Announcements
- 20.3K MPQ General Discussion
- 3K MPQ Tips and Guides
- 2K MPQ Character Discussion
- 171 MPQ Supports Discussion
- 2.5K MPQ Events, Tournaments, and Missions
- 2.8K MPQ Alliances
- 6.3K MPQ Suggestions and Feedback
- 6.2K MPQ Bugs and Technical Issues
- 13.6K Magic: The Gathering - Puzzle Quest
- 504 MtGPQ News & Announcements
- 5.4K MtGPQ General Discussion
- 99 MtGPQ Tips & Guides
- 421 MtGPQ Deck Strategy & Planeswalker Discussion
- 299 MtGPQ Events
- 60 MtGPQ Coalitions
- 1.2K MtGPQ Suggestions & Feedback
- 5.6K MtGPQ Bugs & Technical Issues
- 548 Other 505 Go Inc. Games
- 21 Puzzle Quest: The Legend Returns
- 5 Adventure Gnome
- 6 Word Designer: Country Home
- 381 Other Games
- 142 General Discussion
- 239 Off Topic
- 7 505 Go Inc. Forum Rules
- 7 Forum Rules and Site Announcements