The Ultimate Speed EM Deck

[Marthisdil]

Here's a slightly modified version of the USEMosis that I've been using with great success:

6rk 7ac 7ac 7ac 7ac 7ac 7ae 7ae 7ae 7ae 7ae 7ag 7ag 7ag 7ag 7ag 7ai 7ai 7ai 7ai 7ai 7ai 7ak 7ak 7ap 7ap 7bu 7bu 7bu 7bu 7bu 8pn

(Deck pic attached).

Basically running 1 less frog, 1 less cockatrice, added in 2 jade shields...tossed in a long sword as proof against being decked by other 30 card decks...I've not run detailed analysis, but in AI3 I've only lost 2 games out of about 150 (bad draws are bad draws).  This deck also does well against Top50.

edit: again, didn't keep details - deck wins between turns 5-7 as usual....in top50, if you go against a deck with phase shields, just quit the game lol

M

[spiritkirinblade]

This variant probably slows down USEM's ultimate speed But yeah seems solid, you could try feral bonds if you are going for a longer game

[Marthisdil]

This variant probably slows down USEM's ultimate speed But yeah seems solid, you could try feral bonds if you are going for a longer game

Yeah - not really looking for a longer game as such...more along the lines of a bit of defense...the shields basically shut down firebolt decks (seen in top 50 sometimes)...plus reduces zerg deck damage (if you get the mana fast enough to cast it).  The Mitosis is just knowing when to play on what, but works great against antimatter - put it on the antimattered creature, and it just makes normal creatures....the AI won't ever nuke a creature it's antimattered, even with mitosis on it.

The sword was just a toss in as there are times where the games do get drawn out (rarely) - you technically could just remove it.

[Wolfunit]

He's a robot, don't ask. Just look at the pretty stats and enjoy.

Anyway, now that mitosis + USEM makes the new line of EM mono life *stares at wolfunit* This peaks my interest of a new series of mono life featuring cockatrices + heals + bonds for a new kind of EM deck. Or maybe just a variation of speed heal + mitosis, it will need work.

Good job will, glad you took the time for this.

Wolfie loves you too.

[kirchj33]

Just to reinforce the results willng3 has posted... I've run the comparison 4 times now, but posting this now because its the first time I've run it on low quality...

*USEM=Green, USEMosis=Orange

Games: 200	Avg. TTW: 7.000	Avg. sec/game: 64.055	Avg. Score/Min: 26.91593	Wins: 191	Losses: 9	Ems: 145	Win %: 95.5	EM%: 72.5%
Games: 200	Avg ttw: 7.227	Avg. sec/game: 66.405	Avg. Score/Min: 27.07477	Wins: 194	Losses: 6	Ems:  154	Win %: 97.0	EM%: 77%

Again, more importantly...

Score/hr w/o spins: 1614.9559
Score/hr w/o spins: 1624.486


TSE (http://elementscommunity.org/forum/index.php/topic,21923.msg323873#msg323873) - (rougly equivalent to score w/ spins)
TSE: 1416.002
TSE: 1430.595




Conclusion:

As stated before, this is the 4th time I've run this comparison.  2/4 times USEMosis had a lower TTW, 3/4 times it was more efficient.  I would say this is enough evidence to safely say it is a slightly better farmer than the older version.

In comparing my stats to willng3's... some may be due to a difference in play style, but I would guess the main difference is due to random variation (ex. My draws, opposing draws, elder matchups).  His efficiency is much greater and I would guess this is due to a combination of him having a more powerful CPU (I play on a laptop) and a slight nerf to the speed of Hi AI Speed in the time that has past since he posted his.

[jmdt]

Just to reinforce the results willng3 has posted... I've run the comparison 4 times now, but posting this now because its the first time I've run it on low quality...

*USEM=Green, USEMosis=Orange

Games: 200	Avg. TTW: 7.000	Avg. sec/game: 64.055	Avg. Score/Min: 26.91593	Wins: 191	Losses: 9	Ems: 145	Win %: 95.5	EM%: 72.5%
Games: 200	Avg ttw: 7.227	Avg. sec/game: 66.405	Avg. Score/Min: 27.07477	Wins: 194	Losses: 6	Ems:  154	Win %: 97.0	EM%: 77%

Again, more importantly...

Score/hr w/o spins: 1614.9559
Score/hr w/o spins: 1624.486


TSE (http://elementscommunity.org/forum/index.php/topic,21923.msg323873#msg323873) - (rougly equivalent to score w/ spins)
TSE: 1416.002
TSE: 1430.595




Conclusion:

As stated before, this is the 4th time I've run this comparison.  2/4 times USEMosis had a lower TTW, 3/4 times it was more efficient.  I would say this is enough evidence to safely say it is a slightly better farmer than the older version.

In comparing my stats to willng3's... some may be due to a difference in play style, but I would guess the main difference is due to random variation (ex. My draws, opposing draws, elder matchups).  His efficiency is much greater and I would guess this is due to a combination of him having a more powerful CPU (I play on a laptop) and a slight nerf to the speed of Hi AI Speed in the time that has past since he posted his.

Interestimg as always Kirchj33

[jmdt]

Here's a slightly modified version of the USEMosis that I've been using with great success:

6rk 7ac 7ac 7ac 7ac 7ac 7ae 7ae 7ae 7ae 7ae 7ag 7ag 7ag 7ag 7ag 7ai 7ai 7ai 7ai 7ai 7ai 7ak 7ak 7ap 7ap 7bu 7bu 7bu 7bu 7bu 8pn

(Deck pic attached).

Basically running 1 less frog, 1 less cockatrice, added in 2 jade shields...tossed in a long sword as proof against being decked by other 30 card decks...I've not run detailed analysis, but in AI3 I've only lost 2 games out of about 150 (bad draws are bad draws).  This deck also does well against Top50.

edit: again, didn't keep details - deck wins between turns 5-7 as usual....in top50, if you go against a deck with phase shields, just quit the game lol

M

My 1st version rolled the Jade shield.  Ultimately, they are unnecesicary and slow down the net grinding speed.

[Marthisdil]

jmdt:  I do agree - when going against AI3 they do slow down the grinding speed (as you noted).  Also why I noted that my small changes are also good for grinding Top50 (i like a lil more challenge sometimes when wasting time).  Still averages a good win speed, still EM quite frequently, and can pull out some wins in T50 that you couldn't without the shields.

I 100% agree, if all you want to do is grind AI3, then dropping the shields and going with USEM or USEMosis is the way to go.

[willng3]

In comparing my stats to willng3's... some may be due to a difference in play style, but I would guess the main difference is due to random variation (ex. My draws, opposing draws, elder matchups).  His efficiency is much greater and I would guess this is due to a combination of him having a more powerful CPU (I play on a laptop) and a slight nerf to the speed of Hi AI Speed in the time that has past since he posted his.

I had a bit of discussion about this today and I think I need to throw in my two cents.

I've concluded that a lot, if not all of it has to do with difference in play style/play experience.  The marginal difference between our time per game is significantly greater between USEM and USEMosis which I would assume indicates you put more thought into each of your turns which is either because you needed to make more decisions regarding when and where to use Mitosis or you played more for EMs rather than game speed, with the former seeming to be the more logical explanation here.

As far as actual systematic time explanations goes, I always play on a laptop which means my CPU's power should not be noticeably greater than yours (and even then does not explain why the time difference was so much greater for USEMosis versus USEM especially considering how close the TTW values were between us).  I have honestly never actually noticed this AI high speed nerf which people state took place.  However, those who have noticed it have indicated that the AI high speed nerf took place before I recorded these stats for USEM as well as USEMosis, so that is not the explanation here; I also recorded the stats for USEM prior to USEMosis, so if anything there should be a much greater difference between our stats for USEM as well if this truly were the case.

Given that the results between our stats were not so different for USEM, I can only assume that the random variation for games played while testing for USEMosis were (or I suppose should have been) negligible in this scenario as well.  While I'm not saying this is an impossible explanation, it seems highly unlikely.

Perhaps a better explanation for my reasoning:  When I played Elements on Kong I had a recorded total of ~15,000 victories against AI3 with approximately 2/3 of those I know coming from USEM alone since I used it religiously during my obsessive score grinding phase (during which time I was collecting 1000+ AI3 victories a week for multiple weeks).  I have since done dozens of TTW Studies which I would assume totals approximately 5-6000 additional games versus AI3.  If anything, I would say that this concludes my play experience is greater, and therefore the time taken for each of my actions is substantially less as it is more of a reflex or impulse for each move rather than an actual thought in my case since I know this deck and its opponents far better than most others can claim.  The differences between our statistics are exactly what I was referring to in my USEM versus USEMosis discussion.

[kirchj33]

The differences between our statistics are exactly what I was referring to in my USEM versus USEMosis discussion.

(I felt like I should quote something, but not the whole text)

I am assuming you mean this statement:

"Even though I managed to provide stats which support the fact that USEMosis makes more score/electrum per hour over the original USEM, I would still suggest using the older variation unless you are extremely experienced with AI3 grinding."

Not really sure I need to defend my experience here, but my ~3000ish games of USEM and ~1000ish games of USEMosis definitely pale in comparison the the number you have played.  On the other hand, I am closing in on 400k score and have only played since mid-December, most of this of which has come from AI3 farming as well.

I feel like the actual strategy behind playing USEMosis requires no actual thought in developing after a couple hundred games.  From that point forward, it is all instinctive (meaning you shouldn't encounter anything that actually slows you down to think through). 

While you certainly are the USEM master, I don't think the difference in speed for USEM can be explained by experience or play style, where the deck play itself is mainly point and click.  If we are truly playing with the same CPU capacity, I would say the difference would be that you are faster at maneuvering a mouse than me?  While I don't claim to be the most adept clicker in the world, I doubt I fail to the point of a 2 second difference per game.  I always get the mouse in place and ready to go before my turn hits.  To me, there is obviously something else at play, which I believe still has to do mostly with a hardware difference.  (my laptop is 4 years old?)

In reference to our difference in USEMosis stats, as stated before, this is my 4th time I've run the comparison and USEMosis has performed quite differently on each occasion.  Once, I got the most rediculously efficient stats (sub 7.0 TTW, 83% EM), another time it underperformed USEM.  I chose to post the latest one I did because I felt it was a good representation of the most common performance.  Naturally, the makeup of USEMosis leads to much more variation in games.  There are bad times you are stuck with two mitosis and no creatures, mitosis with not enough quanta to play/activate it, or there are glorious times you start with a bunch of quanta, a creature, and a mitosis to go crazy with while not hitting CC.

All in all, the difference between our posted stats is only a matter of a couple of seconds.  The main point is that I wanted to reaffirm your original assertion that USEMosis is more efficient than USEM, supported by in-depth stats.

[willng3]

All in all, the difference between our posted stats is only a matter of a couple of seconds.  The main point is that I wanted to reaffirm your original assertion that USEMosis is more efficient than USEM, supported by in-depth stats.

Yes and I merely wished to convey an alternate explanation to what you had provided because there were a few explanations left unsolved in your analysis.

Clicking/mouse speed could be a plausible explanation I suppose, but I don't have access to your computer so I can't say, nor would I try to assume because there isn't enough evidence to support it, whether or not your system's speed is less than my own.

Yes, there is a clearly an amount of difference composed by luck when running a set of games for USEMosis, however from my experience there has never been such a significant difference between the time totals from my trial runs.  This is why I said it was possible but unlikely; it goes against what I've seen firsthand.

*And for the record (because I seem to have come across this way in the past), no I'm not saying I don't trust your stats, no I'm not urging other people not to rely on your stats, no I'm not trying to come across as objective or critical of you just because you've posted information which attempts to contradict (even if insignificantly) my data, I'm simply trying to dig for answers because this peaks my curiosity.

[kirchj33]

*And for the record (because I seem to have come across this way in the past), no I'm not saying I don't trust your stats, no I'm not urging other people not to rely on your stats, no I'm not trying to come across as objective or critical of you just because you've posted information which attempts to contradict (even if insignificantly) my data, I'm simply trying to dig for answers because this peaks my curiosity.

Agreed.  I found it very interesting as well, which is why I thought it would be good to post the data.  Thank you for posting this last bit, I did not think you were having any mal-intentions, but simply following your scientific curiosities.  Big-ups.