Recommend
21 
 Thumb up
 Hide
5 Posts

Gears of War: The Board Game» Forums » Strategy

Subject: Gears Theorycrafting rss

Your Tags: Add tags
Popular Tags: [View All]
Scott Yost
United States
Bothell
Washington
flag msg tools
Avatar
mbmbmbmbmb
We picked up this game at Pax Prime and have been consistently getting smoked in 4-player games. I'm trying to crack some of the math in the game and figure out some strategies to improve our game.

First, some of the easier probabilities:

Averages
Expected value (in hits) of one lancer/snub pistol die: 1 (6/6)
EV Locust attack: 0.833 (5/6)
EV Dodge dice: 0.66 (4/6)

% kill with chainsaw (Marcus): 60%
% kill with chainsaw (others): 52%

Expected incoming Wounds (average)
Attacks are undodged and not in cover unless specified.
Wretch (2 locust vs. 1 dodge): 1.15
Drone: 1.95
Drone vs. partial cover: 1.37
Drone vs. full cover): 0.96
Berserker: 3.5

Dodging:
Two dodge dice are worth 8/6 hits on average, and you had to pay one wound to get them. It's a little trickier than that though since you have to worry about overdodging. I wrote a monte carlo simulation for this and I believe that it's a net win to dodge as long as you're down at least two dice compared to the # of incoming locust dice. But it's a VERY small edge - you're only buying yourself 0.25 hits per dodge in the long term.

Berserker
Save a lot of bullets for the berserker. Getting 3 wounds through 3 dodge dice is really hard.
results from monte carlo simulation:
At least 3 wounds with lancer overkill: 41%
At least 3 wounds with grenade: 27%
At least 3 wounds with snub overkill: 17%
At least 3 wounds with lancer (no overkill): 3%

Killing Locust
Most of the "general" AI cards move distant Locust or take away ammo from you, and then draw another card. (the others cause attacks) The specific Locust activations almost all trigger an attack or a spawn.
If you have two of a given Locust around, killing one is always good. I haven't figured out whether killing the last one is worth it - once they run out, their cards become no-ops or spawns. You do save a turn when they spawn, but the spawn cards are rare, so if you've wiped out whole locust groups then almost every AI card becomes a no-op, meaning you'll cycle it very quickly and hit all of the General events again (like Distraction). I suspect that you want to kill the Locust Cs for sure, but that you might end up needing to leave the A spawns and maybe the B spawns at one model and just absorbing the damage. If it takes you two turns or multiple cards to kill the last model and then you cycle the deck it instantly spawns again, you haven't bought yourself much.

Hand Limit
All of the COGs start out at hand capacity. Your heal step is going to be wasted if you don't get some cards out of your hand. Take advantage of guards or follow orders now, since they're basically free. If you have to choose who gets shot or who dodges, prefer a player who has a full hand.

Player Scaling
No matter how many players you have, there is still one locust activation per player activation. Some of the Order cards scale up well - cover fire, for example, can help the COGs draw several cards. The players have more total hit points to share between them, so barring action, they will get more turns to live. But you pay for this by having larger spawns on every board segment, and by paying more life points and turns\locust activations to move an equivalent amount of distance across the board. You can save actions by using Follows, but you're still burning life that you wouldn't have hard to burn with fewer players.
So far we've had difficulty leveraging the orders that scale - my instinct so far says 4-player is the hardest game mode. I'd like to figure out more ways to leverage high player count. We've had fun games at 3-4 player by playing as if it were a 2p game in all other respects. (spawns and locust activations)

The Long Term
We were trying to figure out the general shape of a gears game - is it a controlled descent like BSG where the game is always sapping resources and it's up to you to make them last? Or is it possible by cautious/smart play to overcome the entropy and heal up?
You get two wounds back a turn. You spend one for sure. (ignoring Teamwork, etc.) Every locust activation is very likely going to lead to at least one incoming attack on one COG. As we see above, almost every locust attack is worth about one wound, or more. At those numbers, the players are just breaking even if they draw two cards, use one to act, and then take one hit from the AI activation. So from the get-go, you're already close to breaking even on your health intake.

Ways you get ahead of the curve:
- all locust are far away from you, so they move instead of attacking (COGs only lost one card this turn instead of two)
- no locust are available, so one spawns and ends it's turn (COGs only lost one card this turn instead of two)
- you play an order that draws cards (note - teamwork does not increase your total cards, it just lets you play a more optimal order)

Ways you fall behind the curve:
- any time you take more than one damage
- multiple Locust fire on you
- guarding
- picking up items
- following
- bleeding out, being revived

I think the pace of the game wants to be that there's a heavy firefight and burn when you're outnumbered, and then the COGs are able to catch their breath and restore some health. But in practice, we typically quickly burn down to 0-3 cards each and then struggle to keep our heads above water. I haven't quite figured out what to do about this - there may be better plays we're not making, or maybe it's easier with two players than with four.



Open questions:
When should you interrupt an attack with intent to kill? Given that all incoming attacks are worth at least one expected wound, I think you maybe always want to interrupt. And given that a dodge is only worth .25 wounds, a partial chance to kill is probably better even if you only have 50% of landing the kill.

Comments welcome.

edit 10/8: scaling and how it relates to moving around the board
21 
 Thumb up
1.05
 tip
 Hide
  • [+] Dice rolls
guinch Nudrevil
France
flag msg tools
mbmbmbmbmb
Thanks a lot for your post. We've just lost 5 times in a 3 player game against the berserker, and then 3 times in a 3 player game in Belly of The Beast... I really win more often in 2 players games.

The more the player the harder the game.

Analysing this, I made the same conclusions in various areas. We haven't neither figured out how to nicely manage wounds cards, but the point is to consider the game is harsh from the start of the mission.

Each card is precious, even when you've 6 of them.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Scott Yost
United States
Bothell
Washington
flag msg tools
Avatar
mbmbmbmbmb
My recommendation is to ignore the scaling up based on number of players and use the two player stats. We've found the game to more fun w/ 4 people playing at the 2p difficulty. The game scales up on its own with number of players - I don't see a great reason to additionally spawn a bunch more enemies and stuff.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Basar Cenik
United States
Dallas
Texas
flag msg tools
mbmbmbmbmb
We have played all scenarios at least once, mostly with 2-players, a few with 3-4 players.
My take is that it is very much like BSG, a race to the finish against the board and time. Almost all general events are extemely bad for the players, we think of these as "supercrises" from BSG. The chance to win seems to be inversely correlated with the number of general events you have to endure.

The scaling seems to be bell curve shaped. The reason, I believe, is that the most important "resource" for players is cover spaces. With 1 player, many cover spaces are "wasted" or available for locusts. With 4-players, cover spaces are always in short supply, players have to either slow down or get seperated. 2 players seems to be easiest.
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.