**** Devil Dinosaur (Gigantic Reptile) ****

Unknown

Phaserhawk wrote:

JCTthe3rd81 wrote:

Paintsville wrote:

FaerieMyst wrote:

I know I'm only supposed to play to win and just doing something for fun is a complete waste of my time but I'm going to do it anyway.

At 5 purple, it only costs three. Cheapest board disruptor in the game - plus it's funny. That's win/win in my book.

You are not alone. I am planning on a 4/4/5 build.

Personally, I plan to go 3/5/5. Simply because I will almost always have a better green ability, so that will stay at 3. His red and green are the same, except for the animation. I like the red animation better than green, plus, for such a low Ap cost, it does ok damage. So that leaves purple to get 5 covers.

I want to go 3/5/5, but for the sheer fact that his purple litterally does nothing makes 5/5/3 a no-brainer.

I chose 4/4/5 because he isn't going to ever be high enough damage to be your heavy hitter, so I want to be different. Also, someday, somehow, I see an ability that will complement the purple perfectly and I will be ready.

Unknown

Pwuz_ wrote:

It would be more useful if it was somewhat predictable where it was pulling that black tile from (like always the furthest down or right).

That way you could use it not just for it's top left corner match potential, but in it's ability to swap a known colored tile with whatever random tile happens to occupy the top row.

Perhaps I haven't had enough chance to play with it (especially since mine is simply 0/1/1), but I've yet to see a pattern.

You get roughly 1/7 of a random match 3 for swapping whatever was in the original position in the top left into the random colored tile it selected to do the swap, so at level 5 you get about 6/7 of a match 3. This is a slight overestimation since by this logic Loki's Illusion would do about 4 match 3s on average, but that's because once you start swapping a lot of tiles you end up getting rid of match 3s you just created with additional swaps, which we can ignore for the sake of simplicity and also because Devil Dino's ability does too few swaps to be getting rid of his own match 3s in a meaningful way.

itstime1234

wymtime wrote:

itstime1234 wrote:

Racing up the music charts is the hit single "Where's my dinosaur". Kudos to anyone who gets the reference.

I think you mean Walk the Dinosaur. That is the video I put a link to earlier. http://www.youtube.com/watch?v=zYKupOsaJmk
Whomever originally posted the video did not make it available. UP vot for you for thinking of the same song though.

Actually it was an homage to the Simpsons episode where Bart won an elephant and didn't get it and started screaming where's my elephant in the radio station while other songs played lol

wymtime

itstime1234 wrote:

wymtime wrote:

itstime1234 wrote:

Racing up the music charts is the hit single "Where's my dinosaur". Kudos to anyone who gets the reference.

I think you mean Walk the Dinosaur. That is the video I put a link to earlier. http://www.youtube.com/watch?v=zYKupOsaJmk
Whomever originally posted the video did not make it available. UP vot for you for thinking of the same song though.

Actually it was an homage to the Simpsons episode where Bart won an elephant and didn't get it and started screaming where's my elephant in the radio station while other songs played lol

I remember that episode

FaerieMyst

I've already gotten 9 purple covers. It's like a purple party.

Unknown

FaerieMyst wrote:

I've already gotten 9 purple covers. It's like a purple party.

So you started a second one, right?

Dayv

Phantron wrote:

Pwuz_ wrote:

It would be more useful if it was somewhat predictable where it was pulling that black tile from (like always the furthest down or right).

That way you could use it not just for it's top left corner match potential, but in it's ability to swap a known colored tile with whatever random tile happens to occupy the top row.

Perhaps I haven't had enough chance to play with it (especially since mine is simply 0/1/1), but I've yet to see a pattern.

You get roughly 1/7 of a random match 3 for swapping whatever was in the original position in the top left into the random colored tile it selected to do the swap, so at level 5 you get about 6/7 of a match 3. This is a slight overestimation since by this logic Loki's Illusion would do about 4 match 3s on average, but that's because once you start swapping a lot of tiles you end up getting rid of match 3s you just created with additional swaps, which we can ignore for the sake of simplicity and also because Devil Dino's ability does too few swaps to be getting rid of his own match 3s in a meaningful way.

Do these estimates account for tiles swapped near the edge as well? About 1/3 of possible swaps are right on the edge.

Pwuz_

Sorry if someone else already said this, but:

I was just reading the Digital Times article about the Anniversary and came across this line about Devil Dino.


“We wanted to make Devil Dinosaur play as weird as the character is. He’s a giant dinosaur after all.” Fletcher joked. “He’s going to be running green, red and purple; a different color combination than any other fighter in the game.”

Um...Grey Suit Black Widow?

Unknown

DayvBang wrote:

Phantron wrote:

Pwuz_ wrote:

It would be more useful if it was somewhat predictable where it was pulling that black tile from (like always the furthest down or right).

That way you could use it not just for it's top left corner match potential, but in it's ability to swap a known colored tile with whatever random tile happens to occupy the top row.

Perhaps I haven't had enough chance to play with it (especially since mine is simply 0/1/1), but I've yet to see a pattern.

You get roughly 1/7 of a random match 3 for swapping whatever was in the original position in the top left into the random colored tile it selected to do the swap, so at level 5 you get about 6/7 of a match 3. This is a slight overestimation since by this logic Loki's Illusion would do about 4 match 3s on average, but that's because once you start swapping a lot of tiles you end up getting rid of match 3s you just created with additional swaps, which we can ignore for the sake of simplicity and also because Devil Dino's ability does too few swaps to be getting rid of his own match 3s in a meaningful way.

Do these estimates account for tiles swapped near the edge as well? About 1/3 of possible swaps are right on the edge.

I wrote this up somewhere a few pages back but here's the math again.

For the tile swapped into a random location we'll assume this has 6/49 chance of makiing a match. That is, you've a fixed color so it's 1/49 chance it'll make a match with 2 adjacent tiles (both tiles have to be a specific color and there are 7 colors). There are 6 possible matches for a spot in the middle of the board (2 to left, left + right, 2 to right, 2 to top, top+bottom, 2 to bottom). This overestimates the potential because not every spot has 6 possible matches and you potentially undo a match 3 you with the next swap but trying to calculate all this is a huge pain. So each swap gives you 6/49 chance of making a match 3 from the tile moved from top left to a random spot on the board.

Now for the top left, since all the swap is done at the same time this means only matches are only possible on the rightmost and bottommost section. That is, supposed the top left looks like



But because the swaps are done simlutaneously you know the top left will look like



Although a black tile will be moved to top left, the tile to the right will immediately get swapped with a green so there is no potential for a cascade there. For level 3 you've one match opportunity to the right of green + the 3 below each of those tiles. For level 4/5 you've 2 opportunites to the right of each row and still 3 below, so the total odds are:

level 3: (4+3*6)/49 = 0.45 of a match 3
level 4: (5+4*6)/49 = 0.59 of a match 3
level 5: (5+6*6)/49 = 0.84 of a match 3

Given purple is difficult to cover in 3*, trading 3 purple for about 2 AP of another random color isn't a bad deal, though it's not clear Devil Dino's other abilities are enough to make this investment worthwhile.

Dayv

That's a lot of pretty solid math, but I think you're significantly inflating its odds when you hand-wave away the edge placements. Also, if any of the top-left landing positions are already the final pattern color, you lose one or more of your swaps. I'm tempted to use your math as a starting point and figure out the ratio of odds for corners, edges, and spaces one in from an edge, but I'm supposed to be getting ready for work. My gut feeling is that the 84% chance of a match 3 at five covers would drop below 60% with all variables accounted for... possibly well below it. Anecdotal info from people who've tried it out seems to support this.

All of this also assumes a truly random starting board. If you use this power when you have few matches on the board (which is when it becomes attractive to a lot of players), you're probably going to have less good spots for the tiles to land. To maximize its chances, you should probably use it with a freshly shuffled board, after a shake-up or large cascade.

On the other hand, look at him flailing his tiny little arms!

Unknown

Considering a match chance as 6/49 underestimates the dependent probabilities by about 1%.
If we consider only a center tile X in 5 adjacent tiles. ABXDE

The odds of a match are roughly AB+!ABD+!BDE or B(A|D) + !BDE
1/7 * 1/7 + 6/7 * 1/7 * 1/7 + 6/7*1/7*1/7
= 7/343 + 6/343 + 6/343
= 20/343

For partial side tile: BXDE: D(B|E)
= 1/7(1/7+6/7*1/7)
= 13/343

For an edge tile XDE 1/49 = 7/343

For a given row, odds of the placements are 1/2 center tile (center 4 columns), 1/4 partial side tile, 1/4 side tile (2 side columns).
Overall odds of matching on a row:
= 2/4 * 20/343 + 1/4 * 13/343 + 1/4 * 7/343
= (40 + 13 + 7)/343
= 15/343 ~ 4.4%
Not matching on a row= 1-15/343 = 328/ 343 ~95.6%

Total odds of not matching = (328/343)^2 ~ 91.4%
Actual odds of matching ~ 8.5% ~ closer to 2/25 than 1/7(14.3%) or 6/49(12.2%)

unimatrix

FaerieMyst wrote:

I've already gotten 9 purple covers. It's like a purple party.

Fae. You gotta flay your hands around when you are having a party.. move those purple (pink) hands around..

Unknown

If you want to calculate all the match possibilities we can calculate the bottom half of the board since the upper half is symmetrical. The number of possible match 3 opportunities would be:

4 5 6 6 6 6 5 4
4 5 6 6 6 6 5 4
3 4 5 5 5 5 4 3
2 3 4 4 4 4 3 2

Add all that up and you get 4.5 on average.

So now we get level 5 is (5+6*4.5)/49 = 0.65 of a match 3.

So it turns 3 purple AP -> 2 random AP was closer than I thought.

Unknown

My experience with it is that purple is way more useful than I thought it would be.

Phaserhawk

Intersting math boys. So from what I can gather it's make Dino a bruiser 5/5/3 or a support X/X/5.

Math aside, logic can also help one determine a build.

Lets take Thor, and assume he makes a 3 green match every turn. At the end of his 5th turn he will have enough green AP to cast CotS at the begining of his 6th turn. Assuming tile damage at 61, he will at the end of his 6th turn done 5616 damage to his target which we are assuming is Devil Dino, averaging 936 dmg per turn. Devil Dino has 19180 health, this comes to 20.46 turns or 21 turns to kill Devil Dino with Thor doing this (the extra AP thor has won't be an issue until after this). So at the end of our 21st turn we have to have Thor dead or we die meaning we have to avg 485.7 or 486 dmg per turn to kill him before he kills us. Even if you also matched 3 green every turn.

1.3 (246)
2.6 (492)
3.9 (738)
4. Bite, 2 AP, match 3 5 AP (3496)
5. 8 AP (3742)
6. Bit, 1 AP , match 3 4 AP (6500)
7. 7 (6746)
8. Bite, match 3, 3AP (9504)
9. 6 (9750)
10. 9 (9996)
11. Bite --- Thor Dead

What we see by this analysis is that Devil Dino's strength lies in his ability to absorb lots of damage overtime and eventually killing the opponent by having nearly double the health. So while Dino's purple can by chance take 3 purple AP I got from say making a few match 4's and possibly converting that to some other AP seems to play slightly against Devi's power of just constantly hitting you to wear you down. It is a nice way to burn spare AP assuming you don't have another purple user, but by no means is it a skill that should be attempted to be built around.

Logic can also say, that if one has a better use for green or red than what Devil can offer, than why not opt to make the skill that doesn't conflict with this the strongest, akin to the arguement of not leveling Falcon's purple, because you would rarely cast it anyways, why not make yellow and blue stronger.

While this logic is solid, if one goes x/x/5 it does create a use for Devil, but at the same time removes a strength. To me Devil is going to be my backup plan, he's the tank, the guy that lives when everyone else doesnt, and when he's all that's left because the AI cascaded into Call of the Storm and Rage of the Panther, I'm going to want all the fire power I can get, and that means having Devil at 5/5/3. It's the same reason I want my C.Mags at 3/5/5. Sure the shields are nice, but being able to down an enemy 2.5K sooner or in Devil Dino's case 1K sooner, that's going to win you more matches than flailing arms into random AP

homeinvasion

Phaserhawk wrote:

when he's all that's left because the AI cascaded into Call of the Storm and Rage of the Panther, I'm going to want all the fire power I can get, and that means having Devil at 5/5/3.

Excellent logic and point.

Unknown

If the purple moved everything in a straight line, it would be one of the best powers in the game at 5 covers. Use it 3 times and trade 9 purple for 3 of everything, it like Dino makes rainbow boosts for you.

As he is now, I'll be avoiding him unless I'm bored with time to kill in the season simulator. I can't even Sentry Bomb yet, so taking on Dino is out of the question.

Pwuz_

Perhaps I'm reading your math wrong, but I think the real strength of Prehistoric Arms is not in the tiles moving to the top left. It's strength is moving from the top left to a new position.

So let's say that your top row is:


If there was way to predict which colored tiles were going to be moved, we could potentially get:

Unknown

fakeaseizure wrote:

If the purple moved everything in a straight line, it would be one of the best powers in the game at 5 covers. Use it 3 times and trade 9 purple for 3 of everything, it like Dino makes rainbow boosts for you.

As he is now, I'll be avoiding him unless I'm bored with time to kill in the season simulator. I can't even Sentry Bomb yet, so taking on Dino is out of the question.

Is there something I'm missing in this skill? No matter how the tile swaps are handled you'd never get this scenario to happen. Let's stick with the first swap but it's same for all other. The first time power is used, the top left corner tile is swapped with a random black tile. If you use it again, it takes the top left tile, which is now a black tile, and swap it with a random black tile, which is guaranteed to do absolutely nothing.

Unknown

Pwuz_ wrote:

Perhaps I'm reading your math wrong, but I think the real strength of Prehistoric Arms is not in the tiles moving to the top left. It's strength is moving from the top left to a new position.

So let's say that your top row is:


If there was way to predict which colored tiles were going to be moved, we could potentially get:

How is this scenario even possible? The top row is always going to look like:


after the swaps unless there's something I'm missing about the skill. It'll swap the top left position with a random black tile, then the tile to the right of that with a random red tile, and the tile to the right of that with a random green tile. Since all the previously placed tiles are never the color needed for the next tile to swap you won't have any movement on any of the tiles already placed. Unless a certain color doesn't exist on the board, there should be no possible way to get anything other than the order stated by the ability.