It could be chance. Remember if you have a small sample size, it's possible that (by chance) the placebo control group just happen to get better faster, making it seem like the placebo works 'better'.