Hi Robert,

Thanks for the thoughtful reply.

Examining the ace-v-ace report, i find an interesting observation.

The Fed v. Kli ace v ace raw win% is .51, and the equal player is correcected to .44 +/- .09.

a) the correction factors are large, which at first glance might imply that there is a large remaining skill variance in the ace vs. ace set. And, indeed, player ratings range from 2600-1700 in this set, indicating that at the rails player skill would predict a w/l ratio of about .9. Is it possible to tighten up the ace criteria so the SUPR rating is only a few hundred points?
b) on the other hand, for the individual RPS matchups the MPE error is already so large (+/-0.09) or larger it is hard to make any conclusions about the RPS matchups; so are we at the point were we just don't have enough games yet for that kind of analysis?
c) the composite (all ships) MPE error bar is only about +/- 0.02 to 0.03 for most ships for both the ace and everybody results, that I wonder if a narrower data set could be culled until the error is only 0.03?

-Dave

By Robert Schirmer (Rwschirmer) on Friday, September 09, 2011 - 09:42 pm: Edit

Hi Dave Z,

Yes, I could increase the skill level required for Ace status, but that would shrink the data set and thus increase the error bars on the RPS estimates for individual pairings. Those error bars are, as you point out, already fairly large. Thus what we really need are more matches.

For (c), yes, with the roll-up RPS numbers, you start with smaller error bars, and could keep them relatively small (e.g. 0.02 -> 0.03) even after tightening up the Ace threshold somewhat.

v/r

Robert

By james lee boyce II (Postalpanzer) on Tuesday, September 13, 2011 - 01:56 pm: Edit

It seems the number crunchers are proving the fed is not as much a baseline but more towards the bottom if not the very bottom of the pack.

By Peter D Bakija (Bakija) on Tuesday, September 13, 2011 - 03:45 pm: Edit

James wrote:
>>It seems the number crunchers are proving the fed is not as much a baseline but more towards the bottom if not the very bottom of the pack.>>

Well, no.

If you go back and read the discussion that got us to this point, the data on the Fed indicates that it sits at 12th out of 18 when all players are considered and 10th out of 18 when only "Ace" level players are consider in the calculations.

Which is a little below average, but mostly very middle.

By Peter D Bakija (Bakija) on Monday, October 10, 2011 - 12:17 pm: Edit

Ok, so now between Ithaca Cup (where I did well), Council (where I did not so well), and a few games online (where I'm basically 50-50) in the Neo Tholian, I have a much better sense of how the ship works and what gets it killed.

-Fed: This match up seems super one sided to me. Well, ok, not completely one sided, but, like, 7-3 for the Fed. As I generally observe, the Fed can do fine simply by arming weapons and moving at a moderate speed. Eventually, it will get a shot and blow the Tholian's doors off. I'm yet to come up with a good plan here. I have discovered a few bad plans so far. Brian thinks what works is a couple turns of sniping and using the web to keep the range open is the way to go. I don't disagree with him, I just don't know that I know how to make that work, as the Fed can benefit simply by stumbling into a R8 shot on a rear shield and even average damage will put the Tholian in a hole. At close range, the Fed is going to do significantly more damage than the Tholian. I think making sure that the web is strength 12 when the Fed has to HET into it to get a close shot is key. As is probably a W web. But even then, I think the Fed has a significant leg up here. Which is ok, as the Fed can use a few advantaged games :-)

-Klingon: Haven't played this one yet. The web can nullify drones. The Klingon's good turn mode and UIM and nullify the web. Seems like the Klingon probably has the edge, but no so much that it is a slam dunk.

-Romulan: Tholian has the leg up here, due to plasma. Not hopeless for the Rom, but it is the Tholian's to lose.

-Kzinti: I have played this game *as* the Kzinti many times, and won most of those games. But on paper, the Tholain and snaggle up drones and outshoot the Kzinti.

-Gorn: Same as Romulan, but Gorn will do better. Again, I have played this game as the Gorn a bunch, and usually win. But still, the Tholian has a theoretical edge.

-Hydran: Much less bad than the Fed, as a few volleys of moderate internals can strip the HBs, and the Tholian can generally keep the Hydran from R2 or closer.

-Lyran: The web makes the ESG hard to connect with. The two ships have similar firepower once the UIM is gone. Seems reasonably even, if not slightly in the Tholian's favor.

-Orion: Haven't tried this one yet. I imagine that the Orion has a total leg up with the double HET and reinforcement advantage.

-GBS: This one strikes me as almost as one sided as the Fed. If it just moves moderate speeds and arms guns, it'll eventually win as it is sturdy and shooty. The Kzinti needs the drones more, which are harder to hit with. The Klingon is more fragile and the SP is easy to avoid.

-WAX: In theory, the Tholian has a big leg up here. It is important to play conservatively and make sure that the web is always strength 12 if the WAX can benefit by ramming it. If the Tholian gets tractored (which is not that unlikely if the Tholian zigs instead of zags), it is getting killed.

-ISC: I think the Tholian has a bit of a leg up here. The plasma is easy to snaggle up, and being shorter ranged, will die even after the web drops; the PPD is damaging, but 4x std disruptors over two turns will mostly keep up with a PPD every other turn. I've played it once and my Tholian came out ahead.

I've been stubbornly sticking to a straight web in fights, as I completely irrationally dislike W webs, as they make no sense with what the weapon is supposed to be doing (which I realize is completely irrational...), and it has been working out ok, but I probably need to switch to the W web most of the time.

By William T Wilson (Sheap) on Monday, October 10, 2011 - 04:59 pm: Edit

The W web is just a straight web rotated differently, as represented on a hex grid.

I generally agree with your assessments. One of the reasons I put Tholians, especially the Neo, on the shelf is because of the extreme RPS matchups. Against plasma they always win and against direct fire they are in big trouble.

Snaring drones in web isn't that big of a help because unlike with plasma, the web doesn't kill the drones. What it does is give you the ability to deal with them later instead of now.

By Peter D Bakija (Bakija) on Monday, October 10, 2011 - 07:20 pm: Edit

William wrote:
>>The W web is just a straight web rotated differently, as represented on a hex grid.>>

Well, yes. A straight line that lets you hide behind the corners in it. Still, completely irrational of me.

>>Snaring drones in web isn't that big of a help because unlike with plasma, the web doesn't kill the drones. What it does is give you the ability to deal with them later instead of now.>>

Oh, sure. Usually, I feel very good about weaseling them later. And that often works out ok if you do it well.

By Kerry E Mullan (Nomad17) on Monday, October 10, 2011 - 09:25 pm: Edit

Well I find the NTH does well as with the large number of P3s it can afford to web and then flee using P3s vs drones.

Most of the time the NTH tends to use the web to break off or get unanswered shots before the opponent can then get his shot later. It is a delaying tactic for sure but if you can mount the damage you can come out ahead.

As well the "W" web is much better than the straight as it gives you angles to work with. I agree with most of Peters tactics on RPS except for the GBS and FED.

For both of those 2 opponents maximizing your sniping and then most importantly arranging for a web pass late in the turn(ie phasers after stds were fired earlier) can turn the tide in your favor.

Are either slam dunks? Not really but but using the " Web break" theory it pulls it to 50% range.

Against the FED I actually play to give him the R8 shot on a rear shield after a couple turns of std disr sniping, and if I get lucky a R5 or closer phaser hose(but you have to have the web deny him R5). If he just rolls average you take your 14 internals and turn around to pound him the coming turns(with him being down a shield from your fire).

Oh and to be clear as a web break it's just casting the str 2-4 web closer to you than him so he will not reach the web by the end of the turn(while you will). Just gives you a turn of good firing before he gets to pound you early next turn.

By Troy Williams (Jungletoy) on Tuesday, October 11, 2011 - 06:26 pm: Edit

I added one caveat to your assessment Peter:

"-Lyran: The web makes the ESG hard to connect with. The two ships have similar firepower once the UIM is gone. Seems reasonably even, if not slightly in the Tholian's favor." ESG is not really a factor if you shoot them off while dodging around the web like a scared Klingon.


Great meeting you and losing in a spectacular way to a good Tholian fisting.

By Peter D Bakija (Bakija) on Tuesday, October 11, 2011 - 06:29 pm: Edit

You as well, Troy!

By Mike Johnson (Akira) on Thursday, November 17, 2011 - 03:30 pm: Edit

So what does everyone think about using the web caster as a web fist versus slinging web? Does your opinion change depending on whether we're talking about the Archaeo- or Neo-?

I've started flying Tholian recently and would appreciate thoughts about how best to use the caster.

By Peter D Bakija (Bakija) on Thursday, November 17, 2011 - 04:08 pm: Edit

I've been flying the Neo a lot recently (not very successfully, mind you...), and I rarely use the Fist, as I want to have web on the map, as otherwise, I might as well be a Fed.

If I start a turn in a knife fight, I'll sometimes use the Fist, but even then, if I have a long term plan other than "we both shoot on impulse 1 and somebody wins the game", I'll save it for web.

By David Zimdars (Zimdarsdavid) on Thursday, November 17, 2011 - 06:28 pm: Edit

I think if a Tholian is fighting a drone and disruptor ship which has not deployed drones in a way that can be isolated by the web (or perhaps not at all) then a range 8 overloaded + web fist shot on impulse 25 can do more damage than the opponent can return (potentially). Another way of saying this is that if you are not a Tholian, you should probably do something that make him want to web rather than fist (if you are conservative).

By Peter D Bakija (Bakija) on Thursday, November 17, 2011 - 08:05 pm: Edit

Yeah, to be fair, against Big Plasma, the Tholian can very reasonably use a Fist on the first turn if it can get R8 (see: my game vs Paul last night in the FFAC tournament). The plasma ship can't really afford to launch significant plasma on T1 to keep the Tholian at range--if you launch an enveloper or multiple plasmas, the Tholian webs it and you are out a significant amount of firepower. If you launch a single plasma S, it is probably a pseudo, and even if it isn't, it is going to get shot by a few P3s and taken on a back shield which won't even do internals if it was a real torp, so the Tholian can reliably get to R8 and use the Fist.

By Ken Lin (Old_School) on Saturday, February 04, 2012 - 10:56 am: Edit

PLAY AID: SHUTTLE BAY LIFE EXPECTANCY
Ken Lin
USS Ohio

Similar to a previously published play aid that focused on analyzing expected battery life ("The Duracell Factor"), this analysis uses Monte Carlo spreadsheet simulation methods to analyze the expected survival life of the shuttle bay. The purpose of this play aid is to provide tactical guidance to captains who need to estimate how many internals they can take from a single large volley before starting to lose the shuttle bay.


Table deleted by request. See corrected one below. -- J.Sexton

The way to read this table is as follows. Examine the Klingon line. On average (50th percentile), the 1st shuttle hit will come at the 42th internal. 20% of the time the 1st shuttle hit will come at or before the 35th internal, and 5% of the time the 1st shuttle hit will come at or before the 30th internal.

By David Zimdars (Zimdarsdavid) on Saturday, February 04, 2012 - 05:31 pm: Edit

Cool stuff, Ken. Interesting how Kli is hardly inferior to Fed, for example.

By Jonathan Biggar (Jonb) on Saturday, February 04, 2012 - 07:34 pm: Edit

Question: Does the RKE include armor damage or not? Your table should make note of that. 13 internals (at 5%) to get the first shuttle is pitiful, 23 is bad, but not as bad.

By Matthew Potter (Neonpico) on Saturday, February 04, 2012 - 07:36 pm: Edit

Interesting how the Hydran is in the #4 slot of highest shuttle life expectancy. Actually, their fighters are even higher, since they have alot of padding in that system. Much goodness.

By Ken Lin (Old_School) on Saturday, February 04, 2012 - 08:56 pm: Edit

Jon, no they don't. Add 10 to the RKE numbers to include armor.

By Ken Lin (Old_School) on Sunday, February 05, 2012 - 05:18 pm: Edit

After reviewing, I found an error in the way I set up the spreadsheet, I will post correctred numbers later this week.

By Ken Lin (Old_School) on Sunday, February 05, 2012 - 05:33 pm: Edit

Corrected numbers here. Jean, if you want to go ahead and delete the 1st table, that would probably reduce confusion.

SHIP	5th%	20th%	50th%	80th%	95th%
FED	31	36	41	47	52
KLI	22	27	33	39	45
RFH	34	39	47	56	67
ZIN	37	43	52	63	75
GRN	34	37	42	48	53
NTC	32	38	44	53	62
ORI	13	16	21	27	33
HYD	38	43	50	58	68
ISC, LYR, SEL, WBS	30	36	41	47	53
WAX	41	46	53	60	67
RKE	14	17	23	28	33
RKR	24	29	35	41	47
ATC	25	28	32	38	43
LDR	31	35	42	50	61
JIN	37	43	49	56	62

By Matthew Potter (Neonpico) on Sunday, February 05, 2012 - 08:08 pm: Edit

Woot! The Hydrans went up to (generally) third place! With the 4-shuttle-padding of the fighters, it makes holding the fighters for the classic fusion-style overrun, pretty viable.

By Jonathan Biggar (Jonb) on Sunday, February 05, 2012 - 11:38 pm: Edit

Interesting. I can see why the WAX has the best ranking in the 5% column (all that cargo + center hull) and isn't best at 95% (because it is a small ship).

But why does the KZN start strong and finish even better? It has plenty of Aft Hull and has center warp, but the HYD has 3 more hull, all center.

So why does the HYD not fare better than the KZN in all columns, not just the 5% column?

By David Zimdars (Zimdarsdavid) on Monday, February 06, 2012 - 03:58 pm: Edit

Hey Ken, can you complete the same table for loss of the 4th shuttle box? This is even more important than the start of the loss of the bay, as the last WW won't die until every box is kaput.

By Peter D Bakija (Bakija) on Monday, February 06, 2012 - 05:41 pm: Edit

Yeah, I'm with Dave (although any sort of data like this is always appreciated and welcome :-)

With batteries, knowing when you are likely to start losing them is always important--if you know you are likely to take battery hits on a given volley, you know to use some of them as reinforcement; if you know how much damage you are likely to take on an upcoming volley, and you know you might start losing batteries, you know to pre-allocate to a HET or tractors (or whatever).

With shuttles, generally speaking, what is important is when your *last* shuttle goes up; the first one isn't a big deal. It is knowing when you are likely to lose that last weasel that is really important :-)