|
Simplistically (I couldn't really follow all the statistics), the paper says that:
- the time it takes to do something after moving the camera view in game is a good way to measure reaction time (the 'Looking-Doing Latency'). This very little to do with APM (how fast you can generate input).
- the league a player is in is a good overall measure of skill
- from their data, older players in tend to be slower at this than younger players
- also from their data, age isn't a good predictor of skill
- This suggests that, while reaction times increase with age there are other ways that players can compensate for this - e.g. better use of the game interface, better understanding of how the game works, etc
There was also some other stuff about multi-tasking that I didn't really follow
For a pro player this could suggest that, assuming an equal level of knowledge and proficiency with the game interface (and other non-reaction time factors) a player that is significantly older will be at a disadvantage because they can't react as fast.
In relation to 'only being meaningful if pro player data only is considered' I don't think that would really help achieve their objective of understanding more generally how [males] change as they get older. Pros are, by their nature, outliers to the general population - they are significantly better at certain things than almost everyone else. For example, people who end up as pro players might naturally have faster reaction times than normal people.
|
It seems that the study only shows that the speed of the players tends to decline with age. I think it's very interesting as it corresponds to the age when people start to have a stable life, so maybe the brain becomes less responsive at this time.
Also, the study doesn't seem to measure skill in sc2 but reaction time, so we should not forget that reaction time is just one of the many things that make skill.
anyway, nice work!
|
So true. When I was 22, I was 3 times masters. Nowadays, I'm a washed out diamond player. I should consider retiring on Liquid. 
|
I'll never accept that age is an excuse for being bad at the game or not producing results. You have players in the NFL that are in their late 30's, you have powerlifters in their 40's and 50's. Not to mention people like Jack Lalanne who are able to remain physically active into their 90's until they drop dead. You're telling me people are capable of all that, but can't play games optimally past age 24? Lol.. just lol. Not buying it.
|
On April 12 2014 12:07 Havik_ wrote:
I'll never accept that age is an excuse for being bad at the game or not producing results. You have players in the NFL that are in their late 30's, you have powerlifters in their 40's and 50's. Not to mention people like Jack Lalanne who are able to remain physically active into their 90's until they drop dead. You're telling me people are capable of all that, but can't play games optimally past age 24? Lol.. just lol. Not buying it.
That's not what it's saying. It's talking about averages. If you want to talk about a specific person, you can imagine it like this: suppose there'll be a 45 year old bonjwa in SC2 at some point. That person is winning everything. This study is only suggesting that the same person would have been even better would he have been 25 and with the exact same background of years of playing games and stuff.
|
On April 12 2014 01:18 InvictusRage wrote:Show nested quote +On April 12 2014 01:01 nkr wrote:On April 12 2014 00:26 InvictusRage wrote:On April 12 2014 00:11 nkr wrote:
I would rather read the statistics saying that players above the age of 24 are slightly slower, not that you get slower after the age of 24. There are other factors not taken into account that can give you the same result. This is serious scholarship, guys. Unless, having actually read the article, you can point to a particular factor that they didn't take into account, you shouldn't just assume that they didn't consider it and integrate it into the work. Let me repeat that: this is obviously serious scholarship, published in a peer-reviewed journal. There's a ton of generic "Oh, I dunno about that" in this thread by people who obviously haven't actually read the piece. This isn't some dude saying 'oh, you know, you can just tell that they're slower;' this is a rigorous statistical assessment of a substantial data set. If you're not going to read it, at least give it the credit that peer-review deserves. I've read it, and I don't disagree with their methods or findings, only the conclusion they've drawn. The statistics DO show that the older players are slower, but not that a player gets slower as he/she gets older. If you wanted to measure the impact of age on a persons reaction times, you'd have to gather data from the same set of people over a period of time relevant to the study, and having them maintain the same effort of playing during that period. The statistics simply do not show that players get slower as they get older, because that's not what the gathered data means in my humble opinion. It shows that the older players are slower. There is a significant distinction to be made here, in my eyes. Certainly that's a relevant distinction, but the paper discusses possible other explanations and argues that they are not as likely as age-related decline. You're entirely correct that the data shows that the older players are slower, but the authors also argue that it's not because of any differences (besides age) in the under-24 and over-24 cohorts. For example, they go into detail as to why the slowdown likely isn't because the older players were exposed to RTS games at a later age. But my original post was probably too antagonistic, and picked your post out because it was the more recent when there were definitely much more egregious examples of the problem in the thread, so I should be more careful in the future.
I read that part of the study and in my opinion it was seriously deficient. They didn't put nearly enough work into assessing possible sources of bias and looking at how their methodology or experimental design could be improved to adjust for them. Exhibit A is the factor that SirPinky and countless others have pointed out, namely that older players typically can't devote the time or focus to the game that younger players can. The authors admit earlier in the paper that they observed this in their data ("Older players do report playing fewer hours per week, however (fewer with age; p*<0.001)." However they never mention it in the discussion of cohort effects (which seems to be the social science term for potential sources of bias and unmeasured variables) even though they already admit it exists. Correlation is not causation - to give one example, there is a fairly close correlation between the number of priests in a community and the total liquor consumption (10 points if you can say why).
The statistical methodology is also lacking - it seems to be more like data mining with ex post facto hypothesizing than a real statistical experiment (to be fair, the social sciences are notoriously bad at this, so they may simply be following the normal approach for their field, but that doesn't make it right).
That sounds a bit harsh, so I'll add that I think this is an interesting question and area for research, I agree with the statements about SC2 being particularly suitable for this type of analysis (and their reasons why) and I suspect there is probably a genuine result lurking out there. But I'd find the paper more convincing if it used a more rigorous statistical approach. Lest I be accused of not being constructive, that goes something like this:
1. Decide on the question you're trying to answer.
2. Design an experiment (that is repeatable, predictable and falsifiable) that you believe answers the question.
3. Make a prediction.
4. Perform the experiment and see if you were right.
5. If you were, then repeat with a different data set to confirm repeatability. Also analyse the hell out of any possible sources of bias and hidden variables and try to improve your design to account for them - and get peer reviewers to do the same, preferably including at least one statistician.
6. Repeat iteratively with your design improvements (and new data) and see if the result still holds up.
I'm a mathematician, not a statistician, so I may have parts of this wrong, but I think it's roughly correct. Do all of that and if the effect you're analysing is real (which I think it probably is, but that's only my opinion) then you'll have a much stronger proof.
|
This has been proven with a scientific study in korea based on starcraft progamer Xellos and other gamers using brain scans.
Generally up until the age of ~23 players use a completely different part of their brain based on instinct. Eventually after around this age, this decision making process starts to disappear, no matter what. As a result, response times also slow down.
I think a lot of people are just in denial about this and therefore are highly critical about it, in the end the study came to the correct conclusion regardless of how comprehensive the study was.
The problem is that people are disagreeing with the methodology of the study because they disagree with the answer.
I don't really understand why this study is insufficient. For the question it is trying to answer, I think its sufficient.
|
On April 12 2014 13:03 Chutoro wrote:
Correlation is not causation - to give one example, there is a fairly close correlation between the number of priests in a community and the total liquor consumption (10 points if you can say why).
Ooh, I'll try.
More liquor has more people in rehab/programs has more priests?
|
Interesting how they don't mention the hours of play per week for each age. Starcraft being a game of repetition and time played has a huge correlation to how skilled you are. Funny how studies only show main points they want to "prove" and not state the facts. I'm not saying this study is complete bs, but if the researcher isn't going to look thoroughly at their study. How can you just accept this study. It has a lot of holes. Think for yourself, don't just read the headlines. Most of the time its just someone trying to make a big splash with a pebble.
|
On April 12 2014 07:52 Eliezar wrote:Show nested quote +On April 12 2014 03:24 Chaplin wrote:On April 12 2014 01:24 Eliezar wrote:On April 12 2014 01:06 Chaplin wrote:
I didn't read the study but as long as human history keeps refuting data suggesting something like "too young / too old / unlikely / impossible", I don't need to ^^ This is such a mentally weak statement. This is such a mentally weak statement. On April 12 2014 01:24 Eliezar wrote:
To even suggest that the overwhelming body of common sense and the already established same aspects happening in other sports doesn't exist is akin to proclaiming the world is flat and not round.
Uh, wait! Let me quickly open http://www.dramabutton.com/. The overwhelming body of common sense argued in the (I believe) 70's that runners won't surpass the 100m world record at that time. Suggesting that humans simply can't run faster. (We are now almost half a second faster.) It was also believed and common sense that women had no shot in professional chess (deficit in the visual-spatial realm and all that good jazz). Then Judith Polgar became Grandmaster at a younger age than Bobby Fischer. A whole lot of people still think that you need talent in any given field to achieve expert status. There are a plethora of possible reasons why people in physical demanding sports don't last long (read: physically decline while still being relative young). One might be very well that people train in harmful ways. I can imagine that this even multiplies when you start young. I for one don't question an advancement in sports medicine. (Wikipedia says: "Athletic training has been recognized by the American Medical Association (AMA) as an allied health care profession since 1990." Seems not that long for a profession?) I actually can't believe you are as intellectually challenged as you are pretending to be here, I mean its possible, but its more likely that you are wanting to argue for the sake of arguing and not realizing how completely off base you are being. It is simply a fact that people in every way become weaker over time. Obviously you have the current hard ceiling of telomeres and how many times they can divide. You also have the plethora of studies that show that response time (not reaction time) is slower for humans the older they are and has been looked into with the safety of older drivers. If people's response time slows over time...then yes it happens to gamers too even if Artosis seems to often spout off information that seems to be totally ignorant of that fact. So instead of engaging the study and the other studies that show the problem of slowing response times over age (here is one for you: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2586814/) Your response is to talk about people thinking a record would never be broken for runners? This makes me believe you have to be trolling for several reasons. 1) What does "projections" which cannot be tested but only be theorized have to do with studying the decline the body goes through while aging? Nothing 2) What data about people not being able to continue to perform at peak level do you not believe? It happens in all professions whether mental or athletic. Studies have been done on chess players and how their ability to analyze slows as they get into their 30s. Studies have been done on how older drivers react as fast as younger drivers, but their physical response comes slower. Individuals have been documented as losing a step over time in all major sports leagues. How does people "believing" women couldn't be a grandmaster in chess relate in any way to studies that document that people degrade over time? Literally, your entire post was basically spouting off random dumb stuff that has nothing to do with whether or not the human body decays over time in every way. Your response is akin to me saying...people said you get fat on fast food, but Jared lost weight eating subway therefor StarCraft players won't slow down over time. Again...I have a hard time believing you are actually as challenged as you are portraying yourself.
Don't you find it akward that you have to compensate on a forum? Makes me wonder: "Weak mental state", "challenged"... ever heard of psychological projection?
1) / 2) When is the expiration date on data?
In the time where people generally died in their 30's or 40's, did they also peak in sports at 20-something?
To elaborate a bit: I don't question aging as a factor. I question nailing down a specific age (relating to my first post "too young / too old").
Since you liked Outliers, you could check out Dr. K. Anders Ericsson.
|
Hmm first of all, goob job at writing!
However, biologists found out that people "get slower" at about the age of 29; thats a huge cap from 24. So how can there be such a big difference? The answer lies in day-to-day routine. At the age of 24 many people finish college and go to work. And everybody who has ever worked 40h/a week while managing social duties and a household all at once will agree on the following: They won't have the time needed to be at the same level in starcraft than before.
As many pros already stated out, at some point (high M/GM) it comes down to playing a buttload of games every day in order to get better. And even if you dont have the desire for getting better, it also takes a few hours of gaming to remain the same skill level and remaining the same "Looking-Doing-Latency".
For myself, im soon to be 25, some seasons i am high master, in some i can manage to get in GM league. From my experience at least the reason is NOT that i was older. In some seasons i just have way too much to do in order to just play enough. However if i do have the time to put in the time i want to i dont feel like im playing slower compared to a few years ago.
|
On April 12 2014 17:17 roythereaper wrote:
Interesting how they don't mention the hours of play per week for each age. Starcraft being a game of repetition and time played has a huge correlation to how skilled you are. Funny how studies only show main points they want to "prove" and not state the facts. I'm not saying this study is complete bs, but if the researcher isn't going to look thoroughly at their study. How can you just accept this study. It has a lot of holes. Think for yourself, don't just read the headlines. Most of the time its just someone trying to make a big splash with a pebble.
Look at that. Upset a lurker so much that he makes a TL account just to condemn someone else's hard work. Must be doing something right.
|
On April 12 2014 13:14 sluggaslamoo wrote:
This has been proven with a scientific study in korea based on starcraft progamer Xellos and other gamers using brain scans.
Generally up until the age of ~23 players use a completely different part of their brain based on instinct. Eventually after around this age, this decision making process starts to disappear, no matter what. As a result, response times also slow down.
I think a lot of people are just in denial about this and therefore are highly critical about it, in the end the study came to the correct conclusion regardless of how comprehensive the study was.
The problem is that people are disagreeing with the methodology of the study because they disagree with the answer.
I don't really understand why this study is insufficient. For the question it is trying to answer, I think its sufficient.
Except that it is generally recognized that cognitive decline starts around 29-33 and not 24 (in the sciences). If you take a look at sports where reaction times are paramount (e.g. Baseball), you'll notice that most players peak during their age 27 to 30 years. This is no coincidence as our cognitive abilities peak, and then decline starting at the tail end of this age. It isn't even that large of a drop-off. You don't start seeing large cognitive decline until your late 50s early 60s. There's a reason that you're able to better hit 98 MPH fastballs and 83 MPH change-ups when you're in your late 20's as opposed to your early 20's. If cognitive abilities peaked in your early 20's you would see more Mike Trout's, but you don't. (And please, don't tell me Starcraft is a more plausible case study than a 10+ billion a year industry where reaction times are hundreths of a second for success or failure...)
|
Okay, so finally read it, so a few points of curiosity then my critique, if the OP could address:
So the data was collected by informed consent (online checklist) then they submitted a single replay. I'm curious if there was some stipulation about the replay - could the participants choose any game they wanted to showcase, were they to play a game after the survey and submit it, last game played, etc?
The vast majority come from the United States and Canada - not the best to be making worldwide sweeping conclusions. Were there any Koreans, or enough to have a good sample? I would be interested to see any differences by race/culture
Speaking of race... was the same thing observed for each of the playable races? I'd be interested to see if the same conclusion came from off-racing, too
Due to the repetitive nature of a lot of Starcraft tasks, I wonder if there are earlier-onset injuries - I could see wrist issues and such (typically older users, whether they play a lot of starcraft or not) affecting the LDL as cognitive motor skills
The main problem I had with the study is that it didn't properly(in my opinion - and it seems a lot of amateur research critiquers) take into effect the vastly changing social circumstances that someone 24 and older finds oneself in - specifically in hours played weekly. (well, aside from a completely untouched ("Older players do report playing fewer hours per week, however (fewer with age; p*<0.001").
Especially if you're getting the majority of your data from USA/Canada, the time available for Starcraft will generally go down - and probably by a lot. Because someone is playing less, and probably theorycrafting/keeping up with metagame less, the LDL must go down. So unless you address the hours played (incl hours spent around the game), and work that into the data, I can't buy it.
I also question the usefulness of measuring one single user-decided game for the data. I think also when people play a lot more they will have a larger sample from which to choose for their choice game, so scores will be better. Those who play more are probably more competitive about what they submit, more casual players less. Doing an average of replay stats per participant (including playing all races, beginning and end of daily laddering, etc.) would be better - though I don't know if you have the resources to do something like that.
I like the information on the hotkey variety, and usage, and complex unit information, but I fear they would also be a result of fewer hours/week. You miss the feel of things like how often units spawn, what the perfect balance of unit composition in reaction, the best ways of microing different situations on different maps, etc.
To summarise, if you're going to make sweeping statements that challenge the body science of when the cognitive-motor functions start to deteriorate, you need to look into the total starcraft time / week and factor that in convincingly. Or at the very least add a bolded addendum to your paper that while you feel it's not important, the majority of people who play the game feel that way - so the academics know there's dissenting opinions.
I really liked the paper though, or at least its inclusion of Starcraft II as a legitimate example of cognitive-motor skill and measurement. (EDIT- I also like how the frozen left-pane of the webpage tells me which header I'm on as I scroll down. Technology these days.)
Sorry for the really bad sentence structure, I've been up all night.
|
On April 12 2014 19:31 Wegandi wrote:Show nested quote +On April 12 2014 13:14 sluggaslamoo wrote:
This has been proven with a scientific study in korea based on starcraft progamer Xellos and other gamers using brain scans.
Generally up until the age of ~23 players use a completely different part of their brain based on instinct. Eventually after around this age, this decision making process starts to disappear, no matter what. As a result, response times also slow down.
I think a lot of people are just in denial about this and therefore are highly critical about it, in the end the study came to the correct conclusion regardless of how comprehensive the study was.
The problem is that people are disagreeing with the methodology of the study because they disagree with the answer.
I don't really understand why this study is insufficient. For the question it is trying to answer, I think its sufficient. Except that it is generally recognized that cognitive decline starts around 29-33 and not 24 (in the sciences). If you take a look at sports where reaction times are paramount (e.g. Baseball), you'll notice that most players peak during their age 27 to 30 years. This is no coincidence as our cognitive abilities peak, and then decline starting at the tail end of this age. It isn't even that large of a drop-off. You don't start seeing large cognitive decline until your late 50s early 60s. There's a reason that you're able to better hit 98 MPH fastballs and 83 MPH change-ups when you're in your late 20's as opposed to your early 20's. If cognitive abilities peaked in your early 20's you would see more Mike Trout's, but you don't. (And please, don't tell me Starcraft is a more plausible case study than a 10+ billion a year industry where reaction times are hundreths of a second for success or failure...)
It's possible that hitting a fastball gets easier with experience (being able to recognize the pitchers queues or grip) and actually needing LESS reaction time to be better. I wouldn't be surprised if overall performance would actually be a combination of cognitive-motor performance and experience. If you gain experience faster than you lose cmp your overall performance would continue improving into your late 20s. Once you've reached your skill cap your overall performance begins to decline with cmp decline.
The other thing people are not discussing is the rate of decline. According to the last graph 10 years of decline result in a diamond league player having the "look & react" time of a < 24 platinum league player.
Maybe it's just because I'm already past my 20's I'm more willing to accept that I'm dying :D
|
On April 11 2014 07:55 CrushDog5 wrote:
We find no evidence for the common belief expertise should attenuate domain-specific cognitive decline. Domain-specific response time declines appear to persist regardless of skill level. A second analysis of dual-task performance finds no evidence of a corresponding age-related decline. Finally, an exploratory analyses of other age-related differences suggests that older participants may have been compensating for a loss in response speed through the use of game mechanics that reduce cognitive load.
Dual-task performance could also contribute to increased and improved use of game mechanics that reduce cognitive load. Are you able to make a determination on the possibility that dual-task performance might actually improve as players get older?
|
On April 12 2014 19:31 Wegandi wrote:Show nested quote +On April 12 2014 13:14 sluggaslamoo wrote:
This has been proven with a scientific study in korea based on starcraft progamer Xellos and other gamers using brain scans.
Generally up until the age of ~23 players use a completely different part of their brain based on instinct. Eventually after around this age, this decision making process starts to disappear, no matter what. As a result, response times also slow down.
I think a lot of people are just in denial about this and therefore are highly critical about it, in the end the study came to the correct conclusion regardless of how comprehensive the study was.
The problem is that people are disagreeing with the methodology of the study because they disagree with the answer.
I don't really understand why this study is insufficient. For the question it is trying to answer, I think its sufficient. Except that it is generally recognized that cognitive decline starts around 29-33 and not 24 (in the sciences). If you take a look at sports where reaction times are paramount (e.g. Baseball), you'll notice that most players peak during their age 27 to 30 years. This is no coincidence as our cognitive abilities peak, and then decline starting at the tail end of this age. It isn't even that large of a drop-off. You don't start seeing large cognitive decline until your late 50s early 60s. There's a reason that you're able to better hit 98 MPH fastballs and 83 MPH change-ups when you're in your late 20's as opposed to your early 20's. If cognitive abilities peaked in your early 20's you would see more Mike Trout's, but you don't. (And please, don't tell me Starcraft is a more plausible case study than a 10+ billion a year industry where reaction times are hundreths of a second for success or failure...)
Who said anything about cognitive ability?
Cognitive decline does not equal response time decline. Peak response times do not equal skill either. Especially in baseball, a player often basically has to guess what ball to hit by looking at certain factors such as the pitchers grip and statistics, in which case response time is barely even a factor.
The players are improving inspite of their lower brain response time, this does not mean there is cognitive decline.
This is akin to saying Starcraft players peak at 25 therefore their response times must be at their maximum. No, if Starcraft players peak at a later stage it is inspite of their slower response times. This is because the game isn't purely centered around being the fastest person to react. Its a strategy game.
|
Bogus study.
Age begets stress.
The world of a 35 year old is vastly different from the world of an 18 year old. This study IGNORES pscyho-socio factors of the players. One cannot adequately engage on cognitive tasks when they're, say, distraught for months on end.
The state of the mind is what is key, not the person's age.
|
I have played SC since it came out in 97/98, then BW, and now SC2 since WoL Beta, I have been playing regularly, and in BW I would play 8-12 hours a day , 30-50 games a day, I crunched out 145 games in 1 week last week on SC2 just to get back into masters on time, and I am 26 years old, so I think this study is BS, like people say it has to do with whats going on in your life, and if you can sit and play SC still at an older age.
|
On April 11 2014 08:44 CrushDog5 wrote:
1. You might expect people to take a more relaxed approach, but speed decreases with age. I guess I find it strange that 30 is more relaxed than 25, and 35 is more relaxed that 30 and so on. We have no way to rule it out, but it doesn't seem like to me.
As a 42-year-old who plays Starcraft 2, I'm not necessarily discounting a straightforward slowing of neurological response with age for other reasons, but I can say that the hormonal situation of teenagers and early-20s adults probably favors intensity and focus in practice and gameplay. There's no way that I bring the competitive spirit to the game that I might have at 18. I just don't care that much, and especially for young men, a lot of that is likely driven by their sex hormones. (This is not to say that they view the game as bringing them a sexual advantage of any kind, by the way, it's just that those hormones can induce a higher level of competitiveness and drive in many different types of creative or expressive areas.) This would explain why people at higher ages might be progressively more relaxed, and less focused, about the game.
What would be interesting would be distinguishing between basic neurological factors and higher-level factors like mood and focus. Not sure how you'd construct a study to do that.
|
|
|
|
|
|
EPT
