Thursday, June 22, 2017

A little bit of thievery

Lately I have been thinking about what I could do if I had a normal Blood Bowl team that could steal just one type of unit from any other team.  This includes the number of those units, so normal Skaven can field 4 Gutter Runners, and any team that steals Gutter Runners can similarly have 4 of them on the roster.  What is the best or most interesting thing you could do using this?

The obvious place to start is to pick a team that has glaring, nasty flaws and see if you can prop them up some.  Blood Bowl restricts team to having either players with 4 Agility or 4+ Strength, but you don't get both.  (Barring a single Big Guy).  Adding high Strength players to an already high Strength team probably won't help much because you won't be able to afford them all and by the time you can your team value will be enormous and you will lack Block.  You can surely make a better team this way, but not much better.

The real trick is taking a slow team that is awful at football and making it suddenly a terrifying threat for scoring.  Gutter Runners are certainly the first unit I thought of stealing because they are maximally fast, have 4 Agility, and you get 4 of them.  Dwarves are normally tough and slow but have no quick scoring threats and are unreliable at playing football.  A dwarf team with 13 players including 4 Gutter Runners on it can still field 9 brawling type dwarves on defence and bench 2 Gutters and have plenty of punching power if they want, or switch it up and field all 4 Gutters to create some crazy passing plays if they don't have much time to score.  That team would be terrifying, because they still have the normal slow cage progression tactic available but you have to break that cage FAST or a Gutter will dash in for a touchdown.

On the other side you have teams like the elves or skaven who are great at scoring and have lots of potential for big plays but they have huge problems with getting pushed around.  They also have the struggle that their linemen are fragile and all units need constant replacement.  There are a couple ways you could go there - you could grab Chaos Warriors to have 4 copies of 4 Strength and 9 Armour which would solidify their line immensely, or you could use 6 Saurus instead, which provides absurd amounts of Strength, though it is hard to develop all those Saurus due to them being clumsy.

If you want to go totally nuts you could recruit Ogres and get 6 hitters with Strength 5.  That is the absolute pinnacle of beatdown but does have the huge issue of cost.  At 140k you will only be able to add in 1 Ogre at most to a normal team so it would take a really long time to purchase them all.

When I try to figure out how I would add units to a midrange team like humans I come up short.  They would like both Gutter Runners or Sauruses, as they could develop into a real scoring or bashing team depending on the pick, but no configuration is particularly scary.  To do something awesome with this setup you really want a team that already has one thing they do really well, not a team that is mediocre at everything.

There are actually some teams out there that would make me consider taking linemen on as my choice.  Ogres, for example, are super expensive and you can normally only afford them because they are accompanied by worthless Snotlings.  However, if you grabbed Ork Linemen instead of Snotlings your money problems are worse (they cost 50k instead of 20k) but they actually have 3 Strength so the opponents can't just massacre them effortlessly and they still have 3 Agility so they are capable of playing the ball.  They can't play the ball *well*, mind, but they can play as well as a Snotling and instead of being made of paper they are tough as nails.

On a team like Lizardmen I would be tempted to grab Dark Elf Linemen as my pick.  They are good at playing the ball and have midrange armour so you could play a really serious bashing game with your Sauruses without worrying that every turn one of your squishies is going to die.  You might use a pair of Skinks as dirty players or scoring threats potentially (because your elves can throw the ball) but mostly the team would be just Saurus beatdown and elf football.

In any of these configurations the trick is to make sure you don't commit to more cost than you can afford.  Dwarves are expensive, for example, so swapping some of them out for Gutter Runners or even Elf Linemen is no problem.  You can't just swap out cheap units for expensive ones though, so fixing up teams like Nurgle or Ogres takes more care.

Off all these options I think the dwarves with Gutter Runners scares me the most.  They slot in easily because you can just skip out on buying Dwarven Runners at identical cost, and if you want 4 Gutters to start you just drop 1 Troll Slayer and 1 Lineman to start and you are good to go.  You have a tremendous beatdown game and a sturdy core of players so when your Gutters die you should have spare cash around to buy new ones.  You have a legit quick scoring threat, great caging ability, and a rock solid financial plan.

Monday, June 19, 2017

The stair gambit

When talking about Castles of Mad King Ludwig recently I opined that the Utility Card that gives bonus points for Stairs wasn't so great.  It has substantial potential if you can effectively corner the market on Stairs, but otherwise it isn't exciting.  A game of Castles recently saw me go turbo Stairs to test this theory, and it worked out superbly well giving me a 136 point finish with a wide margin of victory, largely on the back of getting 5 Stairs.


We had Kings Favours for both square footage and number of corridor rooms so the game was ripe for a hardcore Stairs opening.  I started off with the Utility cards that benefit Utility rooms and 200 rooms, and after building 4 sets of stairs I scooped up the Stairs Utility Card!  I would have built more Hallways than I did but one opponent went nuts and scooped up 3 Hallways in a single turn leaving my last Stairs looking lonely and sad.

My board doesn't look that super at first glance but my Utility cards are strong coming in at 10, 8, 6, 6, and 3 points.  I also have 14 points from Stairs and Hallways, and the 200 stack emptied so the fact that I was collecting those was fantastic for me too.  The last two King's Favours were doorways (where I scored 1) and purple rooms (where I scored 0).  However, I cleaned up the two Favours for Corridors and Corridor square footage so I got a solid 17 points from Favours.

One thing that I was wondering was how good Basement rooms are in general.  Clearly if you already have Stairs you might as well build Basement rooms if they look tasty, but are they really worth it?  Do you give up much by simply not having any Stairs at all?  The interconnectedness of the points of various rooms makes this calculation quite challenging, but I am going to have a stab at it on this board.  It won't be easy to generalize it to all games, obviously, so take this as a data point, not a thesis.

I will assign points from cards to the Utility Rooms that generate the cards, not the rooms.  The points from my starting cards will be assigned to the rooms.  I will ignore the 350 Corridor I bought because it was obviously purchased only for King's Favours and isn't representative.

Utility Rooms:  12, 9, 5
Green Room:  5 + 10 coins
Blue Room:  4 + (empty 200 stack = 8 - 3.33 = 4.66)  = 8.66
Yellow Rooms:  4, 4 + (2?) bonus turn(s)
Basement Rooms:  11, 7, 6, 5, 1 + 5 points + bonus turn

My upper rooms generated 6 points + 1.25 coins + .25 bonus turns.
My basement rooms generated 7 points + .2 bonus turns.

The difference between the two set is quite small.  The basement rooms come out on top if you count the kitchen as not completing, but if you treat it as though it completed they are very much on par.

Looking at this board it appears as though my basement and upstairs were similar in scoring on a per room basis.  In that case it appears as though Stairs were a fine investment even aside from the massive stack of bonus points they got me as I did well on basement rooms and having the flexibility to buy them was excellent.  I am not yet convinced that being the third person into Stairs is strong in a game unless the Favours or Utility cards benefit Stairs, but I am eager to see what data other people have in this regard.

Wednesday, June 14, 2017

Cultural Divide

Recently I watched a video by Brian Kibler about the differences in culture between Magic: The Gathering and Hearthstone.  He correctly notes that Hearthstone has a culture focused around the idea that the game is all luck and skill doesn't matter, while MTG has a culture that supports the idea of skill being paramount.

These aren't all or nothing ideas!  Obviously there are plenty of MTG players who whine about getting mana screwed and Hearthstone players who correctly acknowledge that the game has a pretty large skill ceiling.  That said, the trend of Hearthstone players talking about RNG and MTG players talking about skill is real.

Kibler thinks that this is in part because of explicit randomness in a lot of good Hearthstone cards - cards that summon a random minion or make a random spell, for example.  MTG has less of that.  Also the average MTG player plays in tournaments, whether they be small scale Friday Night Magic kind of tournaments or Pro Tour Qualifiers.  Hearthstone players play on ladder and only a tiny percentage take part in tournaments of any sort.

He also thinks that MTG content creators tend to write serious strategy articles while Hearthstone creators make silly decks to play on twitch and youtube and this changes how they are perceived.  While Firebat may bring really tight decks to tournaments he still makes stupid Blood Bloom / Doom decks to play on his stream because those bring more viewers in, so people see top Hearthstone players doing stupid crap all the time and don't see the skill that goes into perfecting and practising a deck.


Kibler's points are right on, but there is more to it, I think.

I remember when I was playing MTG a lot back in Thunder Bay when I was a teenager and it was easy to see that skill was a defining factor.  I won about 25% of the tournaments I entered, and one of my close friends won another 25%.  Mostly anyone else who won was also somebody I knew because generally we were all in the top 8 in nearly every tournament.  When much of your life in a hobby involves tournaments and you see the same people winning every time you really get the impression that skill is the dominant factor.

When I played against good players in top 8 situations I played tight and quiet but it was entirely different in the early rounds against newbies.  If I rolled over someone whose deck really needed some tuning I would often sit after the match and go over their deck to give them pointers.  We would talk about cards they needed, land ratios, what decks other people were playing, etc.  Usually those people would leave some new ideas and also with the definitive impression that I beat them because I was better, not because I got lucky.

In Hearthstone you never get that experience.  When you get beat you just lose and queue up again.  Nobody who crushes you on their way up the ladder sits down with you to say "Hey, you know, your deck could probably use a couple more 1 drops and cut a few expensive dragons."  You don't get people saying "You would have beaten me if you had just Fireballed me in the face on your second last turn."  Lacking those cues it is easy for the player to just rail about getting unlucky and move on.

Hearthstone players also consistently play people of a similar skill level.  The ladder pairs you against people who have won about as much as you, and tournaments are full of top tier players.  You just don't have the newest scrub going into a big tournament, meeting a pro, and getting beat because the pro plays better, at least not nearly as often as it happens in MTG.

When you play against people who are equally skilled, *of course* the victory comes down to RNG.  There is room for individual skillful plays, but on average against a similarly skilled player you would expect somebody's luck to break better and take the victory on that basis.

So while there is an advantage to pitting all the noobs against one another, it does make them think that the game is just RNG based, and the pros end up in the same boat.  Hearthstone pros don't have as much experience at grinding through noobs as the MTG people do, so their games often come down to a single instance of good or bad luck because skill is already controlled for.

This isn't something that needs fixing.  It is just an emergent, accidental property of the way the games are played, marketed, sold, and viewed.  Both companies are doing it right, considering those considerations, but those choices really do affect the way players think about the games they play.

Wednesday, June 7, 2017

Two states

Being 'in combat' is a bizarre and ridiculous thing.  When playing tabletop RPGs like Dungeons and Dragons there is usually a sense of combat time in which characters take turns doing things on a short time scale.  People act sequentially, which is ridiculous from a realism standpoint, but is the only practical way to have tactical combat in this sort of scenario.  It leads to all kinds of weird things - for example, two characters can't walk down a hallway together.  One has to go first, then the other when it is their turn.

Outside this silly (but necessary for our purposes) framework things flow much more naturally in a state I call narrative time.  People can do things simultaneously and even perform activities that don't fit into discrete six second chunks.  You could, for example, give a speech without having to check if every person in the room passes their turn!

The other day Naked Man asked me a rules question that touches on this strange construction.  He wanted to know how to rule it if a player was readying a spell over multiple turns.  If they say "I shoot a Magic Missile at any enemy that walks around that corner" and nobody walks around the corner by their next turn is the spell lost?  Can they just choose to continue readying it?  Can they keep the spell and do something else?  Finally, if they can do all these things, can the players just wander through the dungeon constantly declaring that they are readying spells to attack at all times so the instant they see an enemy they unleash a barrage of magic?

I know what my GMs in highschool games would have done to anyone who tried to ready spells for extended periods like that.  They would have said "oh, rocks fall, you die" and then waited for the player to stop being an ass.  Or they might have had low level spellcasters cast illusions of monsters that walked around corners so the players unleashed fusillades of spells at an illusory beholder.

However, if I want to answer the question of how to handle this in general for a wide swath of players I think I would like to be a bit more thorough and within the rules.

I would definitely allow players to continue to ready a spell against a particular circumstance should it not arise.  If you ready a Magic Missile against an enemy coming around a corner and nobody does, I would say you can just abandon that and do something else next round without losing the spell.

But as soon as anybody says they are readying a Magic Missile during their entire walk through the dungeon, well now that is a different thing.  Readying a spell is a combat action.  It makes no fucking sense outside of the combat time construct, so any time the players are operating in narrative time rather than combat time I would forbid combat actions completely.

If a player said "I am keeping a special watch on that well in the corner in case anything crawls out of it to attack us" I would absolutely take that into consideration and perhaps give them a bonus on a surprise roll or a roll to notice the monster leaping out of the well.  Could even just decide that if a monster does come out they are definitely not surprised.  What I definitely wouldn't do is let them ready a Magic Missile against that eventuality, because they are in narrative time until combat starts.

And since I am the GM in this case, *I* decide when we are in combat time vs. narrative time.

That is kind of a ridiculous solution, but I think it is the only one consistent with the rules of the game.  Combat time is silly but it is an intrinsic part of DnD and you should make use of it when it is helpful, and keeping players from doing silly things with combat actions all the time is exactly the sort of place where you want to enforce the strict duality of game state.

My solution has the nice benefit of feeling elegant (once you accept the combat time construct) and also keeping people from doing abusive things.  Readying spells and then not using them in combat isn't powerful, so there is no need to try to quash it.  After all, you missed an entire turn and did nothing!  All we need to do is prevent players from doing dumb stuff like trying to ready actions for hours on end, and combat vs. narrative time solves that neatly.

Monday, May 29, 2017

The Final Death

In playing Blood Bowl against the computer one of my goals has been to injure their entire team so that they cannot field a single player.  It is not an easy thing to do.  Partly there is the issue of getting all of their players off of the field at all, which is quite a challenge, but even if you achieve that many of their players will be knocked out rather than injured.  They aren't playing right now but they can wake up and come back on!

I played a lot of Blood Bowl 1 and I don't think I ever achieved my goal.  I remember getting a goblin team down to a single player left with fifteen players injured but that last player was stubborn and would only get KOd instead of injured.

I haven't played nearly as much BB2, but today I managed to completely clear the enemy team off the field permanently.  I was playing in my small experimental league with only four teams, trying to find out what would happen if I built a team purely to kill as many opponents as possible.  Would the computer be able to survive in the long run with me in the league?

 There we are.  Eight opponents injured and out of the game, and five more dead, only able to play on as skeletons or zombies.  The opponent was unable to field another player for the rest of the game, and it was only on turn 12 of a 16 turn game!  What a slaughter.  After I scored a touchdown on turn 13 I expected to be able to continue to play, but the game brought up an error message that went like this:

You are unable to field any players.  Both players lose two turns and the opponent scores one touchdown.

What?  It is the enemies that cannot field any players, not me!  Thankfully what happened was the game instantly ended, I was awarded two touchdowns to win 5-0, and things continued on.  I guess they didn't often get to test the error messages that happen when a team is completely injured out of the game so they let this slightly erroneous one slip through.  I was curious what kind of team they would have left after this debacle, and here is what they were going to field for the next game.

They have two players.  Three more are still on the roster but can't play the next game due to injuries, so this team is going up against a 2000 ranked team with only two players.  Naturally they will get tons of inducement money to try to even the score, but even then this is a ludicrous situation.

Also notice the Saurus with 4 Movement Allowance.  He also has a niggling injury, courtesy of yours truly.  The Skink with a penalty to Agility and Armour Value has a niggling injury too!

I checked later on to see what happened in that game, and the Scale of the Sun, despite having only two players, won the game 1-0.  The computer does not take into account team composition when figuring out who wins games between the AIs, it would seem.

So now I know what happens when you play a team of brutal killers in a four team league.  Eventually you absolutely destroy a team and leave them with a collection of horrifically injured players and not nearly enough of those to field a legal team.

For some reason accomplishing this makes me feel good about my life choices.

Thursday, May 25, 2017

Just pass

When I started playing Castles of Mad King Ludwig I thought that passing a turn was a pretty crappy thing to do.  Normally on each turn you spend some cash to buy a room to add to your castle, but if you want you can skip buying to collect 5 coins.  My thought in my first few games was that this was a rubbish choice, one only made by people who mismanaged their money and ended up broke.  I am still convinced that mostly it is taken by people who screw up their money management, but I think it deserved a closer look to figure out if it is a good thing or not.

The key to managing your cash is to realize that if you get really low on funds your opponents can use it to really jam you up.  They can put powerful rooms just barely above the amount you can pay and scoop them up for far less than they should.  They can use basement rooms to put you in a position where you can't buy anything at all and are forced to pass and take 5 coins.  You always want the option to buy things up to 8 cost because then if somebody makes a terrible mistake you can capitalize and nobody can really punish you for being low on cash.  You might want to buy the room at cost 10 or 15 but it is not often going to be such a deal that you get blown out by missing it.

However, if you keep too much money on hand you find out that it is nearly worthless at the end of the game and you might miss out on great upgrades in the interim where you could have bled off excess cash to get more points.

The key is figuring out a ratio of cash to points so you can know when to buy the awesome room at 8 instead of the pretty good room at 2.  How many points more do you need to get for that 6 bucks to be a good choice, given that you have a decent bankroll and aren't worried about going broke?

Some things to think about.  44 room cards are in the deck, generally you go over by 2, and generally 2.5 blue rooms finish, so that leaves about 51 room cards in play.  Mostly 3 of them are left at the end, so that leaves us with 12 rooms per player on average.  Assuming an average score of 110 for good players I think we should model an average purchase giving 9 points.  On each turn where you buy a room you spend 4 coins (I can't defend this mathematically or anything, but a did some figuring and it seems about right for the groups I have played with) and you also give up the option to take 5 coins by passing.  That means that a pretty normal turn where you purchase a room costs you 9 coins and gains you 9 points.

Keep in mind those 9 points are often coming from King's Favour pucks, empty stack bonuses, utility cards, and room completions so the points on the room itself will clearly be far less than 9.

This 1:1 ratio isn't the be all and end all, but it gives us a useful point of comparison.  If you can spend 6 additional coins and get 9 points, it is probably a good exchange.  Next time you are third chair you can skip what is likely a mediocre purchase, take your 5 coins, and you are probably ahead of the game.  If you manage to get more points than the coins you spend you will almost certainly be able to get that money back and be ahead on points in a later turn when your choices happen to be poor.  That is a good rule of thumb!

It also means that if you are staring at a board where you can spend 6 coins to get 3 points it is likely a poor proposition.  Of course if game end is imminent you take the points, but when money still matters you probably don't want to take deals like that because keeping your opponents poor and you rich is important leverage.

One thing all of this analysis ignores is the effect on your opponents.  When you spend money one of your opponents gains money (except when you are master builder, of course) and that matters.  Buying an expensive room because it happens to match your utility cards is good, but if it ships a cheaper room down the line to your opponent that could be a poor choice.  Those are complicated to fit into the basic formulas though, so I have ignored them for now.  I suspect that neither of these things changes the conclusion overmuch.

The real takeaway I have from this is that I really need to consider the cost of skipping 5 coins when buying something.  Buying a garbage room for 1 coin is deceptive as I am actually losing 6 coins to take it.  Usually you will be able to get 1:1 on something on the board but when you don't have a ton of cash you should really think about whether or not to take the money instead.  Paying 4 coins instead of 2 isn't double the price, it is 9 instead of 7.  The absolute differences are the important thing, not the ratio of costs.

I also really need to mind that rooms have a lot of random points attached to them.  Between utility cards, Favour bonuses and empty pile bonuses people accrue roughly 40 points in a game.  That only leaves 70 points for room completions and the actual points on rooms themselves, so when I look at a room and see it is worth 5 points I should really tack on an extra point to account for all the bonuses it might give that I don't know about yet.

So for those looking for a simple set of instructions to figure out how to spend money:  Make sure you don't go so broke that people can take advantage of your poverty, but otherwise just look for deals where you can gain more than 1 point for each additional coin you spend.  1:1 extra purchases are meh, and lower than that is bad unless money doesn't matter anymore.

Tuesday, May 23, 2017

Starting utility

At the start of a game of Castles of Mad King Ludwig every player gets 3 utility cards and must choose 2 of them to keep.  These cards give bonus points at the end of the game based on the castle you have built.  There is a pretty fair range in the quality of the cards, and some are clearly more powerful than others.  There are two questions to answer.  First, which cards are better in general, and how does the existence of the King's Favour change the value of the cards?

Among the utility cards that give you bonus points based on room size there is serious equality:  All of them either give 3 points per room with 6 copies of each room in the game, or they give 2 points per room with 9 copies of the room in the game.  In either case you have 18 points available in the game.  Taking a bunch of copies of a single size does make it easier to get a bunch of points if you empty out that stack, so these are a little better than the 18 points would indicate.

However, the cards that award points based on room colours are much more imbalanced.  When you multiply the card point value by the number of those rooms in the game you get the following amounts:

Purple:  28
Yellow:  27
Blue:  24
Brown:  24
Green:  22
Grey:  20
Orange:  18

This means that colour based utility cards are generally about the same quality as size based cards when you account for the 2 points per depleted stack benefit for stacking size, but the purple and yellow cards seem like clearly the top of the heap.  

There are also cards for round rooms and square rooms, and they both have a value of 15 which makes them quite weak because they are harder to focus on to generate big numbers.  You could potentially combine them with size cards to get a lot of points for each purchase but generally I think they are not good.

The cards that give you points for completed rooms and open entrances seem to cluster really tightly around 4 points per game.  You can get higher, but I don't think I have ever seen them go above 6 points.  They are extremely reliable, require no investment to pay off, and are a fine fallback.

The three utility cards that give points for various corridor cards are really different from each other.  The one that boosts hallways is garbage.  It is extremely difficult to get more than 3 points out of it, and mostly you will get 1.  The one that gives 1 point for every corridor is okay, almost never exciting, and strictly better than the hallway one for some reason...?  The third one that gives 2 points per stairs is tricky.  In the late game it is trash.  On most normal boards it is trash.  But it is possible but quite involved to build tons of stairs and basement rooms completing each other and do really well.  If you empty the stairs stack you rack up 4 points for each of them which is decent.  However, it is a big investment for a not spectacular return so the card overall isn't great.

One card gives you points for every 5 coins you have left at game end.  It isn't great and getting a big score off of it requires you to waste a huge amount of currency.  Mediocre at baseline.

The last two cards give you 8 points for having all 10 room sizes or 7 points for having all 8 room types.  I don't think either of them is great, but I like the 7 point one for all room types much better.  There are at least 9 of every room type and only 6 of each room size for some sizes, so the room type is a much easier condition, plus I find it is actually useful to have all of the room types in your castle whereas having all room sizes has no obvious utility.  Both of those cards suffer from the fact that you can't reliably plan around them in the early game because you don't know what you will lack in the endgame.  They often revolve around desperately hoping you get to buy a room you don't much want on the last turn and often the people aiming for them fail.  I really don't like the idea of a bonus card that you spend resources trying to fulfill and end up scoring 0 points for.  Perhaps the best argument against these all or nothing cards is that they combine terribly with the other cards.  If you want a bunch of blues, for example, it is going to be extremely difficult to also get the variety needed to fulfill one of these cards.

In summary, I think the strategy for keeping starting utility cards can be summarized by a priority list.

Yellow or Purple Colour   2 cards
Other Colour / Size    16 cards
Completed Rooms / Open Entrances   2 cards
Cash/Round Rooms/Square Rooms/Corridors/All Colours/Stairs   6 cards
All Sizes/Hallways   2 cards

Now I want to consider how we combine this with the King's Favour pucks.  Say the King's Favour has the green room puck.  Everyone now has an added incentive to build green rooms.  If you also have the green room utility card, should you keep it?  Generally I see people keeping cards that match the Favours because they think that since they are going to want green cards anyway, might as well score even more for them!

This is not the right approach.  The thing about rooms affected by the Favour is that everyone wants them and there is a real incentive to get at least one of them.  Having a single Favoured room gives you 1 point at worst and often more than that.  Moreover it puts you in contention to jump into first place and score up 8 points if other people get stuck at single Favoured room and you can scoop up a second one.  Even if you just tie with one other player at two Favoured rooms each you grab 6 points, so threatening to do that is powerful.  It is common for one player to corner a lot of a particular colour or size of room under normal circumstances, but doing so when it is the Favoured one is rare and extremely costly.

The key is to remember that nearly every room will be bought at some point.  It isn't as though the ones that aren't worth a ton of points are ones you will just ignore.  You will end up with low value rooms and it will be a lot easier to collect a bunch of them if other people aren't aiming for them particularly, especially if they don't feel that they need at least one.

Ideally you want to take utility cards that give bonuses to the rooms that *aren't* part of the Favours.  You want to be able to cash in on stuff that other people will hand over for cheap, and you don't want to be hunting for room types that everyone wants one of.

So if you see green rooms in the Favours and you have a green room card and a orange room card, keep the orange.  You have a much greater chance of being able to grab a ton of orange and get an outstanding result.

This is especially true because in a four player game you will win more if you take risks.  A player who gets 5 guaranteed points will not win as often as a player who takes a series of coin flips for 10 or 0 points.  You have to have a big score to beat all three players so fighting for the same thing as everyone else and getting predictable, moderate scores is not the ticket.  Aim for something different and try to catch em all.

How much does this matter?  14 of the pucks are coloured ones, and there are 10 other pucks.  (I count the corridor pucks among the 'other' because they work so differently from the coloured pucks.)  When a coloured puck is out that matches a starting card of yours, I would downgrade the card by one tier.  Much of the time this won't matter because even if a blue card is worse when blue pucks are out it is still vastly superior to garbage like the 'collect all' cards or the hallways card.  You will probably still have some other trash you can safely dump but if you happen to have 3 similar cards drop the one that matches the Favour.

Strangely I think the corridor Favours work the opposite way.  Even if they are in play people aren't going to be spending their turns spamming hallways or stairs, it just isn't efficient.  If you happen to have the stairs card or the corridors card when a corridors Favour is out, I think it becomes a really reasonable keep and I would upgrade them both by a tier.  You can pretty easily scoop the Favour and a bunch of points for your cards and that could be quite the coup.

The cash Favour works in favour of the cash card, like I said earlier.  It incentivizes people to complete greens and keep their money around on their Master Builder turn instead of launching it off to the bank.  There will be more money around as people fight for the cash Favour, which helps you get more points off of the cash card even if you don't win the Favour.

The other Favours are broad, so I don't think they have a significant effect on card valuation.

That is quite the wall of text, so I will give a quick summary:  Keep the starting utility cards that give bonuses for room colours (except corridor ones) or sizes.  If you have to choose between those, keep the ones that are different from the Favours.  If you have to choose between lesser cards, keep the ones that match the Favours instead.

*Edit:  I changed the advice regarding the Stairs card because I didn't know the rule that you can't attach Stairs to Stairs, which makes a heavy Stairs strategy a lot harder to do.  It is still possible, but it is much weaker.