29
\$\begingroup\$

I realize that this will depend on a game-to-game, situation-to-situation basis and that this is not a very technical question, but I remember hearing in a tech podcast a few years ago that inflated numbers of points (e.g. 1000 points vs. 1 point) was preferable in games because it was either a standard or because it had been shown to produce better feedback/experience from users. Either case, it seemed to make users happier than lower numbers of points.

I have since been unable to find supportive evidence either way for the theory that inflating scores can keep the user engaged.

Is this inflated-point theory true? Where can I find more information and, particularly, some experimental data on different approaches and their effectiveness?

\$\endgroup\$
14
  • 2
    \$\begingroup\$ This almost seems like a question for skeptics.SE \$\endgroup\$
    – thedaian
    Aug 10, 2011 at 18:28
  • 1
    \$\begingroup\$ That said, there's plenty of games that never/rarely give out under 100 points. Mario Bros, for instance. \$\endgroup\$
    – thedaian
    Aug 10, 2011 at 18:29
  • 3
    \$\begingroup\$ This inflation dates back decades to some of the earliest electrical pinball machines (ipdb.org/showpic.pl?id=91&picno=97 for example has extra zeros fixed in place), so although it's plausible it does increase player satisfaction, I doubt it came out of any rigorous measurement at the time. \$\endgroup\$
    – user744
    Aug 10, 2011 at 18:53
  • 7
    \$\begingroup\$ If you always give out more than 1 point then you don't have to solve the "point" vs "points" semantics problem. \$\endgroup\$
    – Tetrad
    Aug 10, 2011 at 19:06
  • 1
    \$\begingroup\$ I would think that true satisfaction from some number only comes in reference to another number. If a player gets 1 or 2 points from something, getting 100 would probably be pretty satisfying. Or, in RPGs, it's pretty nice to get 10,000 XP from some monster when you remember you were only getting 1 or 2 from slimes in the beginning. Inflation is probably used because most people understand the concept of millions from dealing with money all the time. Hypothetically: what if some person had never heard of numbers before. How would 1,000,000 be satisfactory without some basis for comparison? \$\endgroup\$
    – Bob
    Aug 11, 2011 at 17:24

6 Answers 6

24
\$\begingroup\$

I find myself most satisfied when I see what appear at first glance to be prime numbers.

If the score always just adds a 0 (10, 20, 50, 200, 300 etc.) at the end, I feel cheated.

Multiples of 5 (5, 25, 45, 80, 95 etc.) are a bit better, but you quickly catch on to the fact that your accomplishment is worth 1/5 of what's being shown.

Numbers like 47, 76, 298 etc. feel more meaningful than 45, 75, 300 etc., and you end up with the illusion that there's some really precise math and balancing going on, legitimizing your score.

So to answer your question, big numbers are probably better, but only if it's not obvious that they're inflated.

Oh, and don't forget the thousands separators -- those are achievements in themselves.

\$\endgroup\$
11
\$\begingroup\$

I would say the main reason why some games use such high numbers is, because they don't want to deal with fractions. "0.375 Experience Points" just isn't as catchy as "375 Experience Point". Those numbers have a wider range and variation for balancing without the need of fractions.

The developer can still decide later in a patch to use 1015 points as damage instead of 1000 for the super gun. If the original value would have been a damage of 1, it would become 1.015 damage after the patch. That fractional number just doesn't look so nice rendered highlighted in your HUD

\$\endgroup\$
3
  • 3
    \$\begingroup\$ the online game Runescape made this change. Before your health would be 50 and a typical damage would be 1-3. It led for quite a few '0' damage hits because of rounding. After they essentially increased everything by a power of 10, combat became better. instead of doing 1-3 you did 10-30 (with your health now at 500). \$\endgroup\$
    – user159
    Aug 11, 2011 at 16:58
  • 3
    \$\begingroup\$ This however does not answer the question. \$\endgroup\$ Aug 11, 2011 at 18:00
  • 2
    \$\begingroup\$ @Joe Some 6 years later I have to comment and point out that they did not raise the health by a power of 10 (50^10) but by a factor of 10 (50 * 10). \$\endgroup\$
    – Charanor
    Apr 22, 2017 at 13:04
6
\$\begingroup\$

Preamble: As Patrick Hughes noted, for the "correct" answer to your question, take a sample of test playes. Give one group your game with standard scores, and a second group the same game but with 100x/1000x the scores.

Fractions: Depends on your Game, really. Commenter Joe provided an example of where rounding can occur, but that's about Hit Points rather than scores. It's relatively rare for a game to have adaptive scores, because scores are supposed to reflect the player's skill, whereas, for example, adaptive damage tries to deal more damage to more expierenced players than less expierienced ones. If you suspect fractions for another reason, then avoiding rounding does legitimate for raising everything by a factor of 10 or 100. I can't think of a reason why this could happen, though.

Comparing scores: Again, relatively subjective. Larger scores are easier to compare than smaller ones, though. Try this:

  • Team Alpha got 38 Points
  • Team Beta got 20 Points
  • Team Charlie got 80 Points

versus

  • Team Alpha got 3800 Points
  • Team Beta got 2000 Points
  • Team Charlie got 8000 Points

Which team was better, and how much better? Mathematically seen, both representations are absolutely equivalent.

Satisfaction: I'd say it depends on whether your game is fast-paced or slow-paced. If your game is fast-paced with a lot of action, then "busting the high score" gives me much more satisfication if I cracked the high score not by 80 points, but by 80000. If your game is slow-paced, with lots of tactical thinking, then scores usually aren't the most relevant satisfaction factor at the end of the game. I'd compare the time I needed to beat the round/level. How many mistakes did I make? How many units did I need?

Conclusion: It depends on your game, and know your target audience.

Also, I haven't seen a game which gives you sub-one scores. It's an interesting concept. Getting finaly a score of 1.0319049.

As you can see in the answers and comments, it varies from player to player.

\$\endgroup\$
3
  • 2
    \$\begingroup\$ Your point of fast versus slow games, I like that. I'm going to posit that numbers in the fast games with blindingly high scores function more as sensory overload than they do as useful reports to the player, and the satisfaction comes from the visuals and not the score itself. \$\endgroup\$ Aug 11, 2011 at 22:15
  • \$\begingroup\$ +1 @Patrick Hughes, I was just about to post the same thing. :P It's often used to great effect by displaying points/combos in large text over the play field, rather than just in a simple score counter in the HUD. For more tactical/cerebral games, I would guess that most players prefer something that appears to be accurately judging their skill level, as Rei Miyasaka suggested. \$\endgroup\$
    – mrohlf
    Aug 11, 2011 at 22:53
  • \$\begingroup\$ I don't know about everyone else, but for me once the numbers get large enough I can't really gauge how I'm doing except by comparing the position and value of the left-most 1-2 digits. Smaller numbers are less visceral, but make the player feel that each one has far more value and is easier to conceptualize (see games like Super Crate Box where score is +1 per box) \$\endgroup\$
    – Lunin
    Aug 11, 2011 at 23:35
5
\$\begingroup\$

I'd say higher numbers tend to make players happy. It's better to get $1000 instead of $10, right? Big numbers also go better together with effects. Imagine a particle explosion and sound-effects every time you achieve something in the game. That would certainly look better with an accompanying 14'500, than some low number like 34. Big numbers give the impression of something, well, big, that warrants an audiovisual firework.

Why do big numbers feel like they are more of an achievement? In our daily life, we deal a lot with low numbers. Our daily purchases are usually not in the thousands of dollars, also we don't have thousands of friends or a hundred pair of shoes etc. Numbers from one to ten are probably most common and are very normal to us. It's also the amount we can show using just our fingers. So are numbers from 10 to 50. Higher numbers are more and more rare in our daily life and therefore become special or more valuable.

On the other hand, you shouldn't use numbers for your total score that are so high that your average person doesn't even know how to pronounce it. Anything above several million is too high and too hard to decipher. This is readable: 1'400'000, this is not: 1'400'000'000'000. This means that your score distribution should be in a way that your total score won't go higher than about 6-8 digits.

Beware though, that if you start throwing really high numbers at the user right from the start, it might feel as if there's no room for improvement. A good example are RPGs, where your hero starts really weak and deals low damage at around 1-10, but at a high level with "uber" gear, the damage goes into the thousands. That really gives an impression of progress and achievement.

That being said, it's also a matter game-type and player expectations. As Joe Wreschnig pointed out in his comment, pinball machines started to work with very high scores quite early and this has established as some sort of standard. Doing anything else might dampen the users enjoyment. Eg. "What, only 1'041 points? I got a highscore of 1'300'000 on that other machine".

When you're creating a simulation type-of game, your scores (eg. currency) should be somewhat realistic. Realistically You won't get $1000 per visitor in your theme-park (unless the value of the dollar keeps sinking, but that's another story).

A lot of answers have covered technical reasons for high scores. One that you should keep in mind is screen-size and readability. If your game runs on a mobile device, you simply don't have the space for really high numbers. Also on devices that run on a TV, you're limited by the lowest possible resolution, where readability can become an issue. I'd rather have a very well readable score of 450 than an unreadable 450'000'000.

TL;DR: When designing your scores, think about it as a reward and never forget the context and player expectations. Will 10 points feel good in the context of a pinball game? Probably not. Also try to give the player some feel of progress by increasing scores, maybe giving higher scores more oomph by adding audiovisual effects. So that it really feels rewarding getting better scores.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ It may be true that Americans work with relatively small numbers in our daily lives, but I wonder how this compares with Japanese gamers. It'd be normal for them to spend 15,000 yen at the grocery store, wouldn't it? Perhaps this could be a possible origin to the tradition of using larger numbers. \$\endgroup\$
    – Amplify91
    Jan 23, 2012 at 8:03
3
\$\begingroup\$

Experiment.

Make two games, differing only in the graphics and scoring systems. Give these games to your friends, ask which they liked better.

\$\endgroup\$
9
  • 1
    \$\begingroup\$ Why differing in graphics? \$\endgroup\$
    – Ricket
    Aug 10, 2011 at 18:20
  • \$\begingroup\$ So they don't suspect that you're experimenting on them with scores. It's admittedly a supremely poor experimental design. If you have a dozen friends I can set up a better version. \$\endgroup\$ Aug 10, 2011 at 18:24
  • \$\begingroup\$ Oh give both games to each person. I was thinking of the situation where you take your friends, divide them into two groups, and give one group the normal game and the other group the game with a 1000x score multiplier, and then collect feedback. \$\endgroup\$
    – Ricket
    Aug 10, 2011 at 18:32
  • 6
    \$\begingroup\$ Yet if you're testing the scoring, changing the graphics will render your test results invalid. It's a really hard thing to test, as it's difficult to define how 'fun' something is. \$\endgroup\$ Aug 11, 2011 at 18:12
  • 2
    \$\begingroup\$ A better idea might be to have a control group that plays the game with normal scores which will give you a baseline (try to get as many as possible). Then use a second group (nobody from the control group) with the exact same game, just inflated scores. You would have your answer if you did this enough times and consistently saw that the second group liked the game more. \$\endgroup\$
    – James
    Aug 11, 2011 at 22:41
2
\$\begingroup\$

Indeed higher scores are said to make the player feel like he's playing better, and hence could enhance user experience. This needs to be experimented for you to really find if this makes a difference in your specific game. If you compare an arcade game like Every Extend Extra Extreme and a platformer like Mario, the effect score has on the player is widely different.

As Maik said, higher scores are not only easier on the programming side of things, but are generally much easier for the user to understand at a glance.

Whether you give your player 100 points for a coin, or 2000 doesn't matter: if it is the first object a player picks up in a game, this will be his "score reference". (In this case though I would said 100 points is inherently easier to understand than 2000, but that's more in the realm of psychology).

In my opinion, what's important then is not to necessarily have a high score reference, just an understandable one. What could really give the player satisfaction is when you give him 5000 points instead of the default 200 because he did something well.

A great game for user satisfaction in my opinion is Peggle by PopCap.

In Peggle, when you finish the level, you get a zoom and slow-mo on the last peggle, and huge explosions when you hit it. Very nice. Any bonuses you get multiply your score by huge amounts, again with colourful explosions everywhere.

What is interesting is the mix of visually pleasing feedback, and score multipliers; if I remember correctly, the base score for a peggle is also 100 points, 500 for the ones needed to complete the level. 1 & 5 are mathematically the same, but probably less easy to understand when things are hectic and your score is going up constantly.

So easy, understandable base scores, and huge bonuses and multipliers for doing things well. (Lovely explosions and classical music is also win ^^).

I know that some of the top companies are currently recruiting "Score Designers" to think out how players should be awarded; this shows that it's not something that's simple, and is probably worth investing time into to do right.

\$\endgroup\$
4
  • \$\begingroup\$ -1, restatement of the hypothesis with no evidence. \$\endgroup\$
    – user744
    Aug 11, 2011 at 15:23
  • \$\begingroup\$ @Joe Wreschnig I am clearly not restating the hypothesis, the question asks whether inflating scores makes the user happy, and if you read my answer, I'm saying that inflating scores makes the game more comprehensible, but not necessarily the users happy. I provide more evidence than the other two answers in that I reference to games and my opinions on why they worked. \$\endgroup\$ Aug 11, 2011 at 16:18
  • \$\begingroup\$ "Indeed higher scores tend to make the player feel like he's playing better" - that is not obvious to me and you never defend it with research. It's plausible, but so are counter-hypotheses. Maik's answer makes no such claim but points out a different benefit of large numbers. Patrick makes no claims but describes an experiment that would confirm or fail to confirm the hypothesis. So, they don't need evidence to justify that claim, because they don't make it. \$\endgroup\$
    – user744
    Aug 11, 2011 at 16:25
  • \$\begingroup\$ @Joe Wreschnig I edited to try to reflact that it is my opinion, and that any affirmations would require testing. \$\endgroup\$ Aug 12, 2011 at 13:30

Not the answer you're looking for? Browse other questions tagged .