Winter 2021

Scott Hardie | December 14, 2020
Well, that's not a great way to start off the season.

For the Greg Kelly goo, the same wrong guess kept coming in, from one player after another. Usually when that happens, it's for one of two reasons: Either the photo is of the wrong person, or the clue describes multiple people with no way to tell which is right. But I triple-checked the photo and it was accurate, and the clue doesn't technically describe the other person (he's a frequent guest on that network but not a host as far as I can tell*). I breathed a little easier at not having made a mistake, but I still feel bad for everyone who got it wrong. I certainly did not intend it to be misleading.

I'm deliberately not mentioning the other name here so that it can be a do-over goo, not that most of you don't know the name already.

I worried that nobody was going to solve it, but Chris and Russ got it at the end. Well done, guys. :-)

*There is an article online from September 2020 announcing that this person is going to become the host of his own prime-time news show on this network, and I bet that's the source that players relied on. But I looked at the schedule for that channel and he's nowhere on it, so maybe his show was cancelled or maybe it didn't start yet or maybe it just didn't happen at all.

Erik Bates | December 14, 2020
[hidden by request]

Scott Hardie | February 7, 2021
The "Sci-Fi Namesakes" theme took some twists when I was developing it. Originally it was going to cover all TV genres, but I kept finding so many examples in science fiction that I narrowed it. I originally intended to include the real Stephen Erkel who inspired the name of Steve Urkel (hence part of the reference to "fantastical elements in shows not normally considered sci-fi" in the theme description), but I couldn't find any pictures of Erkel, and besides, it sounds like that guy suffered enough.

Scott Hardie | February 15, 2021
Someone (who can identify themselves if they wish) asked me privately today about game strategy and whether the game encourages everyone to follow the same difficult strategy. I think it's worth responding publicly and asking for feedback from all of you.

This person pointed out that there's a core group of players who all share the same qualities and tend to win: They're committed to playing, they don't make mistakes, they use their lucky cats well, they have some degree of good luck, and they solve each goo at the last minute to maximize their points. The first four qualities are all fair, but the last one imposes a hardship on the players who can't play daily and runs contrary to the spirit of the game. Is it worth changing the rules somehow to disincentivize players from guessing each goo on its final day?

We've talked about this subject in other discussions before. I've pointed out that there are advantages to guessing each goo as soon as possible, from earning additional lucky cats over the course of the season that way, to maximizing the chance that Aka will let you fix a guess that you got wrong. This core group of players don't seem interested in my advice, and they're doing quite well without it, so I think they're on to something.

But today's private message raises two implicit questions: Does it matter whether the last-day advantage is unfair to players outside of the core group who can't play daily and thus can't win, and how would we incentivize the core players to guess earlier if we wanted to do that?

Let's talk about the first question. I'm sorry to single out Richard Slominsky, but his record suggests that something is definitely off about the game rules: He has solved thousands of goos over 13 years of playing and yet has not won the game a single time. And it's not like he's skipping goos along the way like, say, Aaron Shurtleff, who has also solved >1000 goos and hasn't won but who periodically leaves the website. Richard is one of the game's most consistent solvers, literally the seventh of all time and that includes Ruth Cichoski whose newness skews the numbers. (I'll post a list below.) My point is, if Richard keeps playing every goo and solving most of them and has the same luck that everyone else does, shouldn't he have won by now? He's 0 for 53. The difference could well be that he doesn't wait until each goo's last day like other players, and judging from the way that his guesses are clustered (he'll guess a few goos, then disappear for a few days, then guess a few more goos), I'm assuming that he can't play daily or doesn't want to do so.

But the second question is harder: Short of abandoning the pagoda-and-cats system for something else entirely, which I'm not prepared to do at the moment (maybe later in 2021 if I'm convinced by you all), what could we even do about this that wouldn't perpetuate the problem? For example, I could give a small point bonus for guessing earlier, like +12 points for guessing a goo on its first day, +10 points for guessing a goo on its second day, and so on, but that still rewards daily players over less-frequent players. I remember former player Jon Berry dropping out of the game years ago when it switched to a different scoring model that encouraged daily play, because he said he couldn't keep up with playing daily, and ever since then I've tried to find a model that works for everyone, but the pagoda-and-cats system seems to fall short.

So to sum up: Do you think that daily players have an advantage, do you think that advantage is unfair and/or should be eliminated, and if so, how should we go about it? (Bonus question, also raised by the original private message: Does this last-day guessing make the game predictable and boring?)

----------

In researching the statistic about Richard above, I looked up each player's solve rate, which is how many goos they have solved since joining the game, ignoring any goos from before their first solve but not ignoring any gaps in their goo career since then. Here are the top 25:

Russ Wilhelm 95.5%
LaVonne Lemler 86.6%
Samir Mehta 85.7%
Chris Lemler 84.5%
Steve West 82.4%
Ruth Cichoski 68.4%
Richard Slominsky 67.3%
Justin Woods 65.1%
Joanna Woods 55.0%
Matthew Preston 50.5%
Alexis Andino 50.0%
Denise Sawicki 46.0%
Mike Rothstein 45.8%
Erik Bates 37.7%
Steve Dunn 36.2%
Amy Austin 31.5%
Mike Eberhart 25.5%
Aaron Shurtleff 25.2%
Ryan Dunn 16.9%
Elliot Farney 16.6%
Jerry Mathis 16.6%
Tony Peters 16.4%
David Mitzman 16.1%
Scott Horowitz 14.2%
Megan Baxter 12.2%

Samir Mehta | February 15, 2021
[hidden by request]

Aaron Shurtleff | February 15, 2021
I'm like a cicada. I go underground and only emerge every 13 to 17 years...

Obviously, as the example of the person who skips a lot of goos, I am not so much affected by this discussion, so please take my opinions here with the saltiest of grains. With that said:

I am not sure I understand the advantage of solving at the last minute, other than in the context of using the cats well, which I thought was mentioned at the beginning of Scott's post as fair. I don't expect anyone to give away the secret of success, but is there an advantage to the last minute solve that doesn't rely on a cat's ability?

No matter what though, how to fix? Not easy. I recall some time ago (the dark era (error?) of the timed goos) we had a system that timed how long it took to solve the goo. Can we make that a score based process, or something similar? What I mean is, you get points based on how soon after you open the goo for the first time you solve it. This would allow daily players to get points daily as they solved it, but occasional players to solve many goos at once, but still get the similar amount of points based on how quickly they solved it. I don't see an exploit based on daily vs non-daily solvers, but it's a quick thought off the top of my head.

The problem is that I think almost any system can be gamed. Like if we did the betting system above, could there not one day be an optimal number of solves, after which betting will do to get you over the top? I feel like any scoring system needs to, above anything else, reward the solving of goos over everything else (which is probably so obvious that no one else would say it, but I live in the tedium of saying the obvious!)

Answer to the bonus question: There are many games that claim that you can do whatever you want to do (sometimes within reason), but most of them have a "right way" to do it. And, yeah, sometimes that can be boring/annoying, depending on the context. I read an article once about the best way to play Ticket to Ride, and, honestly, it's probably the worst way to play it if you want to have fun, but the claim is it puts you into the best situation to win the game. Is it more important to win or have fun? That's going to depend on the person, I think. For some people, if they don't win, they don't have fun. Not to suggest that the original complainer (who has outed himself) is this kind of person!! To repeat what I said above, whether things change or stay the same affects me not too much, but I think we need to hear from a lot more people before we do anything. (I say "we" like I am going to have any part in any changes that get made! Ha!)

Samir Mehta | February 15, 2021
[hidden by request]

Steve West | February 15, 2021
I like the concept of earlier solve equals higher scoring (e.g. solving on Day 1 equals X, solving on Day 2 equals Y, etc.). The clocks starts upon publication of a Goo not opening same. Timing starting at midnight can garner you a Swift Solution but only qualifies you for a Day 1 solve (Sorry, Samir - you talented bastard).

Waiting to solve later is advantageous with several Lucky Cats (Momoiro, Kiiro, Kuro, and to large extent Beju) give higher rewards the lower you are on the Pagoda. Being in the top spot early in a round rarely pays off because the lead is absolutely temporary because of the aforementioned bigger rewards for lower scoring players and makes Daidaiiro useless. I dread getting Ao and Aka because they're rarely needed. Shiro and Chairo reward equally and can be used instantly.

So, there are times to use right away and times it's advantageous to wait. And there appears to be few negatives to waiting to solve and many times it's positive.

Russ Wilhelm | February 16, 2021
Being a player that waits to put in a guess, I would have to disagree that doing so gives an unfair advantage. I think if that were the case, I'd have more wins, based on the percentages.

So therefore, my lack of winning would have to come down to the remaining elements:
I make mistakes - This is true, ther are times I forget to put in my guess alltogether. But I don't beleive this has affected the outcome of any round to date.
I don't use my cats well - This may have more to do with it. But in order to use them well, the next item has to happen.
I don't have some degree of good luck - Beats me, but I think it's a matter of getting the right cats, at the right time. It doesn't matter how well I play them if I consistantly get Ao and Aka. Right? (I don't, but it's a true statement, based on the percentages of correct guesses to wins)

A better system? Is there one. Maybe, but any system favors someone. Perhaps making it so you can only get any one cat, once per round would make it better. Beats me.

Chris Lemler | February 16, 2021
I have read all the comments and like the ideas from players. However I think the game should stay the same. I don't see any advantage of playing early or playing on the day the goo goes out. I hold back because that is a strategy I have but, I'm not gaining any more points than the next player. I think if this is the case, its taking players strategy away from all of the competitors playing the game.

Scott Hardie | February 20, 2021
The problems that we're discussing with the pagoda and cats system—it has become repetitive and boring, it overly rewards players who play a particular way, it locks in a likely winner before the season is over—are true of every scoring system the game has ever had. It's the eternal curse of the goo game. I'd be happy to change to another system, since changing every couple of years at least keeps the game feeling fresh, but the last system went so poorly that I'm reluctant to change to a totally different system unless I'm really, really confident that the next system is going to be fun.

I was thinking about Jeopardy! and how it has barely changed its format over the course of decades and yet remains interesting. I think part of the appeal is new players: Imagine if the same 10 people played Jeopardy! year in and year out; it would get boring fast. But really, the appeal of Jeopardy! is in *watching* the game on TV, not in *playing* the game, which would be a nerve-wracking nightmare even without a national television audience. It's a poor model for the goo game. The same could be said of professional sports.

Timed goos to end the season wasn't a terrible system. It just became repetitive, and too rewarding to players who played a particular way (fast research and guessing without total confidence in the answer). Long before that, we had what was essentially the opposite system, a series of extremely difficult goos to end the season until all but one player was eliminated, which was also repetitive and too rewarding to one kind of player (the methodical and meticulous researcher), among other problems (it was a drain on me to make so many hard goos in a row). The badges system was an attempt to make players stretch their skills a bit by rewarding lots of different kinds of ways of playing the game, and I still love that principle behind the system, but it had too many other problems.

I have a bit of a damned-if-I-do, damned-if-I-don't feeling about changing the goo game again. Every time I change to a new scoring system, I risk running off players who liked the old way or who feel like they can't win the new way, and there's no corresponding increase in other kinds of players to make up for it. But if I leave us on the same system indefinitely, it gets boring (you're 100% right Samir) and there's an inevitable attrition of players that way. The answer is of course to promote the site and attract more players, something that has been quite problematic in the past. So this subject is a downer for me, but it's important to discuss.

It occurs to me that if the goals of most goo scoring systems are 1) to reward every single goo solved, 2) to randomize the outcome a little so that it's unpredictable and anyone might win, and 3) to avoid someone getting a lock on victory before the end of the season, then the purest approach to this would be to have a lottery or raffle system: Each goo solved gets you a ticket. After 90 days, a ticket is selected at random. If it's one of yours, you win. This system would *not* be popular. :-P

Samir, I like your ideas. The gambling idea is close to something we did already, but it does add a sense of player agency into the randomness element; you choose how much risk you want to take on. I really like the idea of adding more variation into the cats' rewards based on the progression of the season; that has terrific potential to be a game-changer late in the season. I also like the idea of you having to do something to activate your cat; it's not just luck that earns you the reward but some kind of specific action that you have to take, whether it's a simple action (create a goo, solve a goo) or scaled to the size of the accomplishment (something like X points for solving a goo in 1000 - X seconds).

I really appreciate all of the feedback so far! Every comment has given me a lot to think about. If other people have ideas, please share them. I'm going to ponder these a little bit longer and make a decision soon.

Regarding Ticket to Ride, yes, there's a perfect route in the original game that will win you the game ~90% of the time: Vancouver > Los Angeles > Santa Fe > Oklahoma City > St. Louis, with a branch down to Atlanta and another up to Chicago & New York. If you complete that route, you can just keep drawing tickets for the rest of the game, because nearly any combination of three tickets has at least one ticket that is either already finished or only needs a quick connection to a nearby city like Boston to finish. The publisher caught on to people employing this strategy and changed it in subsequent releases like the 1910 edition that have a much better balance in the ticket cards. It is a bit boring to play this way, and unfair to people who don't know the trick; I've only tried it once against human opponents but they were expert gamers who likely had heard of it. (I deservingly lost that game because I misread the box; we were playing 1910 and I thought we were playing original. Oops.)

Aaron Shurtleff | February 20, 2021
Oh gosh, Scott. That's not even the way I was talking about! In the article I read, it advocated that if you made it your goal to create all the routes into certain cities, it will make it so that only you can complete a lot of routes. It was essentially win more by not allowing your opponents to do anything than it was doing stuff to give yourself victory. There was a bit more to it, but the important part was denying your opponents any way to score points to win. To me, that is winning at the cost of being a butt (and I am pretty sure that, at least in TtR: Europe, they tried to prevent this, using train stations I think they were called). It is like playing Monopoly to just get one one every color set and deny everyone monopolies and hope you win through pure luck. Yeah, it can be done, but it's the worst! And of course, that is only my opinion. There are lots of people who think winning that way is perfectly fine, and certainly within the rules of the game. It's just not my cup of tea.

Scott Hardie | February 20, 2021
Ha! That's such a simple and mean strategy that I guess I'm pleased with myself for not being evil enough to think of it before. :-) I haven't seen people play that way to my recollection, but I do know people who will choose and complete the smallest tickets they can at the start of the game, then rapidly run out the proverbial clock by making only 5-car and 6-car connections as fast as possible to force the game to end before other players are ready.

My gaming group plays a lot of Terraforming Mars, and we've all grown wary of the one occasional guest who employs the city grid strategy, building cities spaced evenly apart on the map so that no one else can make them and so that you get all of the victory points from other players' greenery tiles. I'm a believer in natural consequences; play like a jerk if you want but don't expect other people ever to help you, let alone invite you back.

Chris Lemler | February 20, 2021
Hey I brought an idea to Scott to add a new twist to the current game. How would you all feel about a "whammy" type situation. If you guess the goo correct, you collect those points for difficulty. If you guess wrong you get those points subtracted from your total score. Would like everybody's thoughts please.

Scott Hardie | February 22, 2021
As I said to Chris privately, my thoughts on the "whammy" idea are that it's good for introducing more variance in scores and for allowing players to lower their scores for strategic reasons, but bad for further enriching players who are already ahead. It's similar to what Samir suggested above about reintroducing wagers into the game for more variety. I'm open to more discussion on it.

I've been kicking around some ideas for an overhaul to the lucky cat system that could solve many of the problems mentioned in this discussion. I'd appreciate some feedback on these ideas as I consider which, if any, to implement. Here's what I'm considering, numbered to make them easier to discuss:

1) The pagoda system would stay the same, except that point values would be lower and progress with the season: You would get 1 point for solving a goo from the first week of the season, 2 for a goo from the second week, and so on, so that players who fell behind would have a better chance of catching up late in the season.

2) I would replace the old cats with entirely new cats. This would thematically justify them having new abilities. (Some abilities would return, but I have lots of ideas for new abilities.)

3) Instead of one chance to earn a cat's help every night at midnight, you would instead get a chance every time you solve a goo, and every time one of your created goos is published. (In either case, let's call this the "triggering goo" for sake of discussion.)

4) Each cat would have a favorite topic, and would have a much higher chance of being the cat that helps you if the triggering goo is from their favorite topic. (Example: "This cat loves knowledge! Solve a Science or Literature goo to have a higher chance of attracting him.") This way, players can try to target the specific cat that they want most.

5) You wouldn't "have" a cat with exclusivity. You would still have four days to accept a cat's offer to help you, but the same cat might also offer to help other players during that time, and you might get offers from multiple cats at once. However, a cat will not make a second offer to you while it still has one pending.

6) Each cat would have three different abilities that could help you. You would get to choose which one of the three to use. This way, hopefully none of the cats would feel as useless as Ao or Aka.

7) Each of the three abilities must be earned by doing a different task. When the cat first offers to help you, which ability can be earned by which task is randomized, so you can't get too comfortable just doing the same over and over. The three tasks are a "Feat of Knowledge," guessing every current goo; a "Feat of Inspiration," creating a goo; and a "Feat of Compassion," donating 5% of your points to another player. Not all of those options will appeal to everybody, but you can always pick a different task to do of course, and besides the whole cat system is optional anyway.

8) I would publicize exactly how the odds of getting a cat change, so that players could more directly manage their odds. So far I have kept it ambiguous with statements like "being lower on the pagoda increases your chances of getting a cat" without making the numbers clear. In addition to explaining the numbers, I would make each player's numbers explicit on the page so that you can actually see yours.

9) Your chances of getting a cat would increase with the newness of each goo that you solve; ie. solving a goo on its first day would improve your chances more than solving a goo on its last day, with a sliding scale in between. To me, this is less about creating incentive to solve earlier than it is about offering players a strategic choice: Do you solve earlier to get more cats, or solve later to get more benefit from each cat? You might find yourself going both ways at different times.

10) The math for various cat abilities would not be based on your numerical level on the pagoda, but on your relative score. For example, say player #1 on the pagoda was in the lead with 100 points, player #2 had 95 points, and player #3 was way behind with 30 points. Right now, the old cats treat the gap between 1-2 and the gap between 2-3 the same. The new cats would give an even bigger benefit to player #3 because of the larger gap in points between them.

I have ideas written out for all of the new cats' abilities, including wagering/whammies and other new possibilities such as blocking point steals and extending a goo's expiration date to have more time. For simplicity's sake, I won't get into all of them right now (I think the above is enough to digest), but I can share what I'm thinking if people want to know. Most old cat abilities are repurposed in the new system, so you'll recognize some of them.

At this point, my biggest concerns about the above changes are: Can I get them done in time for the next season? (Probably.) And, are they so complicated that they take over a game that's supposed to be about guessing celebrities? (Possibly.) The lucky cats are our game's equivalent to the random items in Mario Kart; they're fun and more variation among them is good, but they shouldn't completely take over the game; the driving should remain the primary focus.

I'll continue considering all of this as I collect your feedback and make a decision soon. Thanks in advance!

Steve West | February 22, 2021
This sounds fantastic! I have no clue to strategy. Can't wait to try it and develop one.

Scott Hardie | February 22, 2021
Thanks, Steve!

I thought an example might help:
- You solve a music goo.
- A notice appears on the page that you have attracted the music-loving cat.
- She offers to give you help in one of three ways: You can get an instant 25-point bonus, but you have to guess every current goo. You can block other players from stealing your points for a week, but you have a create a goo. Or you can guess again on a current goo that you guessed wrong, but you have to donate 5% of your points to another player.
- You create a goo.
- You return to the cats page and click on the button to activate the music-loving cat's second ability. Your wish is granted; she disappears.
- Later in the season, the next time she appears, she'll have the same three abilities, but what you have to do to earn each one is different.

Performing the feat is only good for one cat, and does not automatically trigger a cat's ability; you have to go click a button to get your reward. This way, if you have three cats, simply creating a goo won't automatically trigger the "create a goo" reward on all three simultaneously. You can choose which cat to apply it towards, leaving the other two inactivated. If you want the "create a goo" reward on another cat, you'll have to create a second goo.

Samir Mehta | February 22, 2021
[hidden by request]

Erik Bates | February 22, 2021
[hidden by request]

Scott Hardie | February 23, 2021
Thanks for the feedback! I'm still thinking about it all.

Happy 23rd anniversary to Celebrity Goo Game today! I'm involved with this game daily and even I still can't believe it's still going, and still inspiring the kind of discussion that we're having right now. I'm glad that you all still care about the game enough to keep playing and to make it better. Thanks for sticking around. :-)

Scott Hardie | February 27, 2021
The good news is, I have decided to implement all of the above. I think all of those changes will be for the better, and I'm glad to hear support across the board.

The bad news is, I won't be able to implement it in time for the Spring 2021 season tomorrow. I'll keep working on it over the next three months and debut it with the Summer 2021 season.

If you have more suggestions, ideas, comments (pro or con), or anything else that you want to share, please continue to do so. I appreciate all of it. :-) Thanks!

Scott Hardie | July 29, 2021
After re-reading the entire above conversation in preparation to finally code these overdue lucky-cat changes, I can't help but suspect that I was way overthinking the problem. I wrote:

The goals of most goo scoring systems are 1) to reward every single goo solved, 2) to randomize the outcome a little so that it's unpredictable and anyone might win, and 3) to avoid someone getting a lock on victory before the end of the season.

Samir Mehta | July 29, 2021
[hidden by request]

Steve West | July 29, 2021
I can echo those sentiments. I contemplated skipping a season because I dread playing with those Lucky Cats. Get Shiro enough times and you're almost a lock to win. I like the sound of these changes.

Scott Hardie | August 6, 2021
Thanks for the feedback! I haven't heard from anybody else about this proposal, so it sounds like the consensus reaction is positive. I have made all of these changes and they're ready to launch on the weekend of August 28 between seasons.

You'll be able to see your own random bonus points per goo, but only on the last day of the season will you see everyone else's to find out who won.

It sounds like any change at this point will be a good one. Here's hoping for the best with this! :-)