Michael Link
msg tools
Avatar
mbmbmbmbmb
I'm now about 40% of the way through my attempt at all start worlds vs. all start worlds using Brink of War. After posting a win rate over 50% on hard difficulty, I began pondering whether there was something about BoW that made it easier for the player or whether a strong gameplan would carry the day using The Gathering Storm only.

I should mention that I was introduced to Race by a friend who bought the base game plus all 3 expansions, sight unseen, right away to save money on shipping! As a result, I ramped up to BoW for multiplayer within about 30 total plays, and had played less than 10 games of solo with TGS only before adding expansions. And those were all on easy, basically serving to get me over the initial learning curve before continuing onward in Rebel v. Imperium and then BoW, quickly adjusting to playing on Hard difficulty only.

So, inspired (and frightened), by Neil Thomson's solo series, I decided to re-enact his experiment and downgrade to TGS only for 50 games on Hard. Here's the results:

Record: 13-37 (26%)

Playing As:
Doomed World: 2-2
Damaged Alien Factory: 1-1
Epsilon Eridani: 2-4
Separatist Colony: 2-4
Old Earth: 2-5
New Sparta: 2-5
Earth’s Lost Colony: 2-6
Alpha Centauri: 0-3
Ancient Race: 0-7

Robot Playing As:
Doomed World: 7-0
Damaged Alien Factory: 6-0
Epsilon Eridani: 4-0
Old Earth: 6-1
Separatist Colony: 4-1
New Sparta: 3-2
Earth’s Lost Colony: 3-2
Alpha Centauri: 3-3
Ancient Race: 1-4

High Score: 62 (as Doomed World)*
Low Score: 11 (as Old Earth)
Highest Losing Score: 61 (as Earth’s Lost Colony)
Lowest Winning Score: 29 (as Separatist Colony)

Robot High Score: 73 (twice, as DAF and ELC)
Robot Low Score: 24 (as Earth’s Lost Colony)
Robot Highest Losing Score: 62 (as Old Earth)*
Robot Low Winning Score: 34 (as Alpha Centauri)

*This was a controversial decision that originally favored the robot, but on further review I had failed to award myself 1 VP chip in an earlier round. This resulted in a 62-62 tie, which I won 7-6 on tiebreaker. This was the only time the tiebreaker was needed.

Record by Number of 6-Devs Robot Claimed:
0: N/A
1: 4-6
2: 7-12
3: 2-11
4: 0-7
5: 0-1

Expected Record on Medium: 22-27-1*
Expected Record on Easy: 31-17-2*

Score adjustments for expected records:
For Medium: -3 for each of first 1-2 6-Devs and all beyond 3rd, -7 for 3rd
For Easy: -3 for 1st, then -7 for 2nd and 3rd, -3 for all others

*The rationale for this is that if the robot could not have played a 9 point dev chit, he would have had to mill. Sometimes this would whiff or draw a 0, and sometimes it would hit a 6-dev anyway for 6 points. On average, I guesstimated that this would be worth 2 points for the robot, so a difference of 7 points. My record-keeping was not detailed enough to determine what the tiebreaker outcome would have been, so ties are simply ties.
___________________________________________

The Post-Mortem:

Wow. So much for even coming close to beating the robot on Hard with TGS only, I got pretty much schooled. Interestingly, my projected win rate on Easy in quite close to what the robot's was on Hard, and Medium would have been nearly a dead heat. Also, it is somewhat eerie how similar the rank order list of home worlds is for both me and the robot--we both valued the start worlds nearly identically! Both of those findings, however, probably have more to do with chance than anything else. Some more salient take-home points that deserve serious mention:

1. The Windfall Start Worlds Sucked. That is an incredible deviation to what I usually find in my live games, where it is usally a huge advantage to start with a windfall world. On the human side, it is undoubtedly due to the inability to do a lot of trade-produce (or possibly that I SHOULD have been doing more trade-produce and just sucking up the economy growth for the robot). On the robot side it is less clear. With Ancient Race, the fact that it starts with only 3 cards and 1 credit (read: significantly nerfed) is huge. With Alpha Centauri, it tends to place a dev chit right away and then spend the game at zero credit. Its robot action being settle, but without a ton of big worlds to hit and relatively small card draw, is also pretty weak.

2. Produce-Consume Was Quite Prominent. I frequently found myself trying to "go P/C" much more so than I do in BoW. While I realize that P/C was overall much stronger in the base game, it is often harder to do in solo due to the robot's intrinsic power there. Nonetheless, it can be so hard to overcome the 9-point dev chits that getting something like a 6p2c engine by round 3 was something I found myself often shooting for from the outset of the game. The problem comes in when the robot is unexpectedly able to get the economy up to 3 or 4 before you get there (see my last session report to get an idea of what this looks like).

3. You Are Hosed Without Early Synergy. The robot is so strong because its economy is straight-up abstracted and 6-Dev synergy is straight-up abstracted. It therefore is not fettered by having to find interlocking cards and the cardflow to get it down. It takes a good deal of synergy to crank a 6-Dev up to 9 points in TGS only--I find that somewhat easier with BoW. In BGG-documented victories over the Hard Robot TGS-only, early synergy was a prominent factor, such as:
A. morningstar's logs (examples: New Sparta + 2/2 uplift in starting hand, get ImpLog, tack on Terraforming Guild. Also stuff like getting down every Alien card, drawing playable military windfall on 1st round, etc. Multiple windfalls + ImpLog + Terraforming Guild was common theme)
B. MrWeasely's log (started with Inv Credits and TLOTUG as New Sparta, then got multiple useless robot actions)
C. Erak's log (big Alien synergy, 3 interlocking 6-Devs)

4. Even the Best Have Failed (Best = Logged on BGG). Quite little is documented in the TGS forums about people's exploits with multi-game series vs. the robot on Hard, but there are a few:
A. NeilThomson (4-46, 8%)
B. rbelikov (8-21, 27%)
C. GerryRailBaron (32-40 doing all start worlds vs. all start worlds on medium. By his adjustment, if the games had been on Hard, would have been 5-66-1 (7%))
D. fnord, under Rindel's thread (66% on Easy, no documentation of Hard)

This was about all I could find documented, and my results appear to be on par with these. Compared with rbelikov, who has the highest win rate documented within the forums that I could find, I was pretty close--I was 7-22 after 29 games. My point in posting this is that I wish there were more session reports and solo series published, and also that I'm not sure how much better can be done on Hard. My estimates on highest theoretical win rate is around 70% on Easy, 55% on Medium, and 35% on Hard. Those are purely estimates. I'd love to hear about anyone doing better, and how they achieved it!
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Rus
United States
Mountain View
California
flag msg tools
Avatar
mbmbmbmbmb
theright555J wrote:
Compared with rbelikov, who has the highest win rate documented within the forums that I could find, I was pretty close--I was 7-22 after 29 games [...] My estimates on highest theoretical win rate is around 70% on Easy, 55% on Medium, and 35% on Hard. Those are purely estimates. I'd love to hear about anyone doing better, and how they achieved it!


Very interesting! While it's probably correct that the robot got easier with the expansions, I seriously doubt 35% is the highest theoretical win rate for hard on TgS. My 8/29 (28%) TgS hard rate is from a time when I was learning the game, and I am much better now. Furthermore, while I am quite good (if I do say so myself ), I am by no means top tier. I suspect that the theoretical maximum for TgS on hard is well over 50%.

What I do find in my (multi-player) games, however, is that now that I am used to BoW, I don't do as well in the base game, because my brain now carries BoW strategy inertia that just doesn't work in the base game. I suspect your win rate in TgS may suffer from a similar effect and that the expansions don't make the robot as much easier as people think.

Edit: my TgS rate on medium is 11/18 (61%), already higher than your proposed 55% maximum. These are small number statistics of course and variance may play a role, but again, most of these 18 games were among my first several dozen games (counting multiplayer and solitaire), and I couldn't have reached the theoretical maximum that quickly!

Incidentally, I think BoW hard theoretical maximum is probably well over 90%. I myself am at 70% (44/63), and in many cases my losses were merely due to carelessness and impatience. (I was actually at 90% for my first 10 games back when I wasn't as rusty as I am now Race these days is far and in-between unfortunately!)
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.