That's the way I would summarize. When I lived in Las Vegas for 24 years in the sports betting world all the sharp guys were maximizing positions by jumping on the early numbers in projection where they would end up. For example, let's say they wanted 10 times ($11,000 to win $10,000) on the favorite -3. But they thought that favorite was too low and would probably rise to -4, due to injuries or public opinion..whatever. In that case they would wager much more than 10 dimes on -3. They might bet 50 dimes. They do that because there is immense value in an NFL game of giving -3 and taking +4 on the same game. They would win if the outcome landed on 3 or 4. Plus when you lock in that immense wager on -3 and the game indeed bumps to -4, that enables countless favorable options like jockeying money lines in relation to the spread or pointspread propositions, etc.
In that market you quickly learn that scalping (arbitraging) and middling are more profitable than merely grinding out winners. The first time I played a middle in a basketball game and it landed it was an incredible rush. That's a 20x payoff because you win both sides. I had $550 on -3.5 and $550 on +5 in some obscure college game involving San Diego State. When it finished on exactly 4 I won $1000. My risk was only $50...if the outcome finished on one side or the other I would win $500 from one wager but lose $550 on the other one.
Realities like that are the reason it is much easier to make a living on sports betting than conventional wisdom allows. Outsiders seem to envision people making a robotic wager on one team or another and then waiting for the outcome like a dolt. Hardly. It is constantly jockeying toward favorable math.
In my early political wagering years in Las Vegas I was frankly playing head to head against suckers. There was no legal political wagering in Nevada. The loudmouth sports wagerers in Las Vegas were overwhelmingly right wing blowhards. This was in the '90s. They were not nearly as sharp on politics as sports and I often took advantage. The Republican slant was in their mind alone and it often did not reflect reality. For example, a few months after the '94 midterm I had one guy give me Even money on Bill Clinton being reelected in 1996 for a huge bet. It was ridiculously favorable from my standpoint. The guy assumed '94 would repeat. He had no clue toward situational variables like an incumbent whose party has been in power only one term, the scenario Clinton enjoyed.
I also was in 16 man betting pools every two years, mostly with guys who tilted overly right.
Once fixed political odds showed up on offshore websites, the odds were quite primitive. For example, if a candidate had a 3 point edge in polling the odds reflected a 3 point sports advantage, like -165 favorite. That was a joke. A 3 point edge in polling is multiple times more meaningful than 3 point power rating edge in football. I was betting on the favorite time and again, no matter which party it was. Eventually sharper guys entered those offshore sites and drove the prices up, but not until Nate Silver became prominent in 2008 did political wagering dramatically change on the fixed odds pricing. In 2006 I got Lieberman to defeat Lamont at -160 (5/8) in the Connecticut general election. Meanwhile, if Nate Silver had already been established with percentages on his site I guarantee those odds would have been closer to -1000 (1/10).
I hope I am answering your question satisfactorily. I realize it is not quantum or brief. I thought it was best to provide an overall rundown of what I have experienced over 25+ years. Predicit is most interesting because of all the markets offered and also the comments below each market. Bettors will describe what they are thinking. Those odds in general seem to be a blend of the 538 number along with other political odds sites like Predictwise and Optimus https://0ptimus.decisiondeskhq.com/ , etc. That makes sense because sports odds are the same concept...a blend of the reliable mathematical power ratings.
The political odds on Predictit shift every day based on polling. That's why I say it is reactive. Check any race and if you see a 3 cent shift or 8 cent shift then invariably it is due to a new poll. Heitkamp, for example, was trading in the 40s in North Dakota but last I checked it was down to 23 cents after those back to back polls with a double digit deficit.
The reaction is measured based on the slant. Let's say North Dakota were a Democratic state but polls similarly showed a huge jump for Heitkamp's opponent. In that scenario, the odds would not have moved nearly as much because there would have been a ton of skepticism. When the poll change is in the direction of the ideological slant, and the direction people have expected the race to turn, then the price can move sharply and never come back. Political gamblers don't mind investing 75 cents to win a quarter. Lopsided political races are exponentially more reliable than let's say a 10 point favorite in a football game.
I like to wager 70% on Republicans right now because my pet category is that liberal/conservative breakdown by state. A very frequent occurrence recently is Democrats trying to win a state with 40+% conservatives. It very rarely happens, unless that Democrat is an incumbent. Obviously I hope Beto wins in Texas but that state has 44% conservatives and 3/4 of them say they are very conservative or extremely conservative.
My theory is that the polling does not adequately account for the ideology of the state, and that rigid liberal/conservative breakdown. If a Democrat trails by a few points in a state with only let's say 35% conservatives, that's not an impossible task. But bump it up to 44% conservatives and with a Democrat trailing by a few points, and that is a completely different ball game. Yet the odds seldom fully account for the difference. That's where I look for advantage, taking advantage of the rigid ideology. Let's say Beto trailed Cruz by identical polling but we were talking about Florida, not Texas. Florida has 36% conservatives. There is no way in hell I'd be giving 65 cents on Cruz. I'd be looking to wager on Beto, since a comeback would be significantly more available.
= new reply since forum marked as read
outlook.
for a period for which I never gave much attention to measuring it's length.
and an outside force comes along and 
pumps some positive vibes
