Gil Hova's game Prime Time is an economic engine game whose primary mechanic is drafting. Players can draft special power (network) cards, actors, television shows (which require different combinations of actors) and turn order/income bonus for the next round. The drafting itself is straightforward. The strategy of the game comes in deciding which shows to target. Shows have varying effects on your income and points (viewers). You can see a full explanation of the rules in the video above. The version of the game being tested was from some time ago. You'll see sidebars where Gil will respond to the feedback and provide updates on how the game is now.
Gil says:
Michael has playtested Prime Time a lot over the years, and he’s seen the various twists and turns the game has taken. It was completely different when he started testing it. I believe the only thing it has in common with the original prototype is the drafting nature of the game. Everything else is completely different; in fact, it used to be an auction game!
In fact, the video Michael refers to is from six months ago, at least; the game rules have changed considerably since then, both from his feedback, and from feedback from others. It’s a much better game now!
There have been a bunch of changes to the game since then, like Ads, which should have been in the game all along. Ads increase a Show’s income, while Stars reduce a Show’s income but increase its Viewers (VP). There’s also an excellent mechanism where a Show or an Ad will be better or worse depending on some condition it must meet when you are attaching it to a Show. That’s nowhere in the version you’re seeing in the video, but it’s been testing very well.
This change was tricky for many reasons. I generally like changes that streamline and simplify my game. This change did the opposite; it made the game more complex, so much so that it added about 15-30 minutes to its play time, and makes the game no longer playable with 6 players. But it also improves the game enough that I feel okay with that.
A part of being a game designer is needing to kill your darlings sometimes. The random Budget Chip draw was a darling I had for a long time that I finally killed very recently. It’s a really cool mechanism that actually brought nothing to the table. I now have one set of chips for the entire game (except for the first Season, which gets a different, more gentle set of Budget Chips), which reduces rules density, component count, and swings that are out of the player’s control, which players hated in a tight planning/optimization game like this. It was tough to do, because I loved that mechanism, but the game plays so much better without it.
That change wasn’t enough, so I made a real nerve-wracking modification: I made the last Star or Ad on each Show optional. This change really concerned me, because it goes completely against the game’s core tenet of knowing exactly how close each player is to acquiring a Show. So far, that change has played out extremely well. The lower cost threshold means that there’s more incentive to replace a show, but the optional nature of the final Star/Ad means shows might be more rewarding if you’re prepared for them. And losing the transparency of how close players were to acquiring Shows didn’t turn out to be such a huge deal, because players still have plenty of information to work with.
There are a ton of other small changes, like that turn order is now re-calculated between turns by players’ scores, so the player in last place goes first in turn order, and the player in first place goes last in turn order. This handicapping mechanism doesn’t work for all games, but since Prime Time doesn’t have any significant hidden endgame scoring, it works really well here.
Also, Stars are also no longer split across gender lines, which I’m really happy with; that was an annoyingly false distinction. Instead of male and female Stars, you now have Stars and Ads. That improves the game both thematically and mechanically, which turned out to be a no-brainer.
The immediate thing that struck me after the playtest was that there was an imbalance between the choices between points and money. Players could build their shows using parts which maximized income or using parts which maximized points. In most games with this points versus resources dilemma, the key to winning is to find the point at which you switch from gaining resources to spending down those resources to gain points. For excellent examples of these types of games, look at St. Petersburg or Dominion. In both of these, going for points too early will cripple your ability to purchase more efficient point-scoring cards later on (save for the unfortunately unbalanced . But waiting to long to make the switch will allow your opponents to scoop up all of the points cards before you can act.
The exception that proves the rule.
Image from BoardGameGeek.com
However, the tradeoffs as given on the show cards and actor tiles made going for points the overwhelmingly good choice, even in the early game. Recognizing this on the first turn let me run away with it. Luckily for the designer, this type of numerical imbalance is easily taken care of with iterative testing and does not require a new mechanic or clever rule change to fix.
Gil says:
And sure enough, at this point, runaway leaders in the game are not nearly as common as they used to be.
You may wonder why I say “not nearly as common”, instead of saying “there are no runaway leaders, ever”. I feel that in a skill-based game like this, there should be some possibility of a runaway leader. Completely eliminating runaway leader in a skill-based game would have to mean that players’ performances are not as meaningful, especially in the early- to mid-game. If a player performs significantly better than everyone else, she should be rewarded. I prefer that to a game that’s so nerfed, my choices are meaningless, but hey, at least the scores came out close!
At one point, I suggested that the designer might use an auction mechanic to solve the tradeoff. It would even be thematic to have network executives bidding over actors and scripts. However, the designer had tried this approach in an earlier version and it was not successful. In addition, auctions are sometimes used as a crutch to auto-balance games. The designer might not have wanted to rely on that crutch.
Gil says:
Yep, he remembers! As I said, this was once an auction game, but the auction mechanism clashed pretty harshly with the rest of the game. It was one of the first darlings I killed. Doing that was tough; the game spent a year on my shelf because of it! But once I removed the auction, the game worked a million times better.
Therefore my suggestion was that he should come up with a specific cost model that relates points to money. In this version of the prototype, no specific cost model was used to create the various attributes on cards and tiles. This allowed me to spot tiles that were so imbalanced towards points that it was the optimal strategy.
Creating an explicit cost model, where a tile that gains you X points should be considered equal to one that gains you Y money at the same resource cost, will allow the designer to more effectively perform iterative testing on the various ways to combine X and Y. Without it, the best he could achieve is spotting problematic tiles one at a time. Having a cost model that dictates all tiles (with tiny deviations allowed), will reach the "correct" set of tiles in far fewer iterations.
Collectible card games (the decent ones, anyway) make sure to employ this design methodology. Instead of coming up with hundreds of cards individually costed, the design team assigns costs to various attributes and abilities of each card. This way, when they design a card, they are much less likely to make a costing mistake which creates categorically better and categorically worse cards. You can always nudge it from the prescribed amount, but you'll be in the ballpark. I had firsthand experience with this while on the development team for a now-defunct CCG.
Gil says:
Michael’s point about point modeling is huge. It is something critical to do in an economic game, and shame on me for not doing it early enough. I’ve crunched the numbers and valued each turn at roughly 3 points, although the exact value will depend on situational factors like what round of the game the players are in, and how far in the round they are. This has helped me weight a bunch of elements in the game, like Shows, Stars, and Network Cards.
The player who was randomly set as the starting player was given less money to start, which is a common and useful tactic to balance out turn order. However, that player's starting money was so low compared to what was needed to take advantage of going first that his first round was significantly worse than any other player's. The goal with this type of balancing should be to negate first-mover advantage, not to create first-mover disadvantage.
The best way to approach this type of turn-order handicap is to figure out what resources a player absolutely needs to start the game in a productive manner. Then, instead of taking away from that number for the first player, add to it for the successive players. This method ensures that players late in the turn order feel compensated without the earlier players feeling as if they were unable to play the game at all.
Gil says:
In a game with such a huge significance of turn order, I had to pay a lot of attention to the first round’s dynamics. In the current version of the game, with Michael’s feedback, the first player has enough money for 2-3 turns in the first round, while the last players have enough money for about 4-5 turns. This turns out to offset the initial random turn order advantages, and by the time the second round starts, positions have been scrambled nicely.
Unfortunately, this method is made more difficult to implement by a separate design choice made. The deck of shows if fully randomized. Since the shows can vary highly in terms of both acquisition cost and stars required (and those stars have acquisition costs), it is very difficult to determine what the minimum starting money should be for this game.
I'll admit that a fully-randomized show deck adds variety to the experience. However, I don't think that variety is worth the disruption to the game's economic curve. A semi-randomized deck, with shows that are cheaper or require few or no stars in the first half and the larger shows in the second half would accomplish two goals. First, ensure a smoother experience for the start of the game. Second, it would allow the game to reach a dramatic crescendo with high-value, high cost shows appearing at the climax of the game.
Gil says:
Michael has advocated for a stacked Show deck, but I disagree. I want each season to have a mix of good Shows and bad Shows. If all the Shows that came out in the final Season were good, that would make turn order in the final Season less meaningful. I want turn order to always be tight in the game.
So instead, I played with Show values until I had Shows that were equally effective and interesting in both the second and fifth (final) Seasons. I have a separate set of “starter” Shows that come out in the first Season; that seems to take care of starting turn order imbalance. The fully randomized Show deck now works great; playtesting backs me up on this.
There’s a similar thing with the Network Cards. There are some cards that just aren’t interesting if they come out in Season 5. I really detest the idea of stacking the deck for this as well, because almost every other card works at any point in the game. So I may just dangle some Viewers on those cards for people who have them at the end of the game.
Gil also had several mechanics that encouraged players to replace shows. Each player only has three slots for shows and placing a new show in an occupied slot discards the underlying show. Each show's value in points generated changes between rounds (representing a growing/shrinking audience for that show over time) and there are action cards which accelerate or decelerate that change.
However, a few show cards were designed to have a high cost with a high payoff at the end of their life cycles. Unfortunately, this meant that the only reliable way to receive that payoff would be to acquire that show in the second round (of five), when players would likely not have the economic engine needed to acquire it. Also, the disincentivize players from replacing old shows with new ones, which is supposed to be the fun part of the game. Dedicating 1/3 of your play area to not changing for 4/5 of the game is not enjoyable.
These high-cost/high-payoff shows should have their growth curve shifted to be in tune with when they are likely to be played. Otherwise, they might as well not exist. Perhaps the designer wants the option to be there. However, making players commit to a do-nothing strategy at the start of the game risks making the game seem boring to that player (and others, if that strategy ends up winning).
Gil says:
There are no more shows with high payoffs at the end of their life cycles; instead, they all arc by their second or third seasons. By the final season, they have all “jumped the shark” and score their lowest. This incentivizes players to replace them, which is always an interesting choice, and as Michael notes, much more engaging than doing nothing. Additionally, no show ever has more than four seasons’ variance.
Michael played one game where he only ever acquired three shows. His feedback on the game at that time was that “it was [expletive] boring”. I’ve since found that I want players to acquire 5-6 shows per game. I’ve reduced the Star requirements for each Show to make this more attainable. The average number of Stars (and Ads) a Show needs has gone down from 2 to 1.5, and then to 1, with the optional Star/Ad I mentioned earlier.
The game manages to do a good job of mechanically representing its theme. There seems to be nothing fundamentally wrong with the overall structure. The changes required to this version of the prototype were not of a particularly difficult nature.
Gil says:
I’m really happy with how well Prime Time’s mechanisms match its theme. And as a game, it’s gotten to a point where it’s a ton of agonizing fun. I’ve spent a long time tweaking the game; it’s finally starting to show it.
Thanks so much to Gil Hova for giving us a look into the long-term development of Prime Time. If you are a game designer and would like your game to be featured on this series, please contact me.