In the exemplary embodiments, the confidence client 124 may act as a client in a client-server relationship and may be a software, hardware, and/or firmware based application capable of processing the polling data from the smart devices 110 via the network 108. The confidence client 124 may be configured to determine a confidence score for the polling data that is received from the smart device 110. The confidence score may be a measure of an accuracy of the state of the world or the device ecosystem that may be determined from the polling data. That is, the confidence score may represent a computation based on the polling data from each of the smart devices 110 that identifies how confident the agent 120 is that the preview of the state based on the corresponding smart device 110 is a good representation. For example, when the polling data is related to a cloud service to track health conditions of a user utilizing the smart device 110, the simulation from the corresponding smart device 110 may be used to generate polling data with a high likelihood of representing all of the effects to be produced as other data from the other smart devices 110 may not be relevant. Therefore, the confidence client 124 may determine a high confidence score for such polling data (e.g., a score of 0.95 where the confidence scores range from 0 to 1). In another example, when the polling data is related to a cloud service in which a plurality of users utilizing a plurality of different smart devices 110 contribute in maintaining data, the data available on one of the smart devices 110 may only provide a partial picture. Only through consideration of the data from each of the smart devices 110 utilizing the cloud service may an entire picture be determined. In such a scenario, the confidence client 124 may determine a relatively low confidence score for each polling data on an individual determination. For example, the confidence client 124 may determine that the confidence score is a proportion that the smart device 110 associated with a given polling data is a part of the cloud service. In a particular example, the confidence client 124 may determine a confidence score having a value that is an even distribution among the smart devices 110 utilizing a cloud service where a device ecosystem including four smart devices 110 will each have a confidence score of 0.25. However, those skilled in the art will understand that there may be a plurality of other considerations that may affect the confidence score in a device ecosystem involving two or more smart devices 110 (e.g., priority level, contribution level, a measure of how current information is used in generating the polling data, etc.).